-

The cell culture laboratory

CAROLINE B. WIGLEY

1. Introduction-animal cell culture: past,

present, and future
The history of animal cell culture dates from the turn of the twentleth century (fable /), with Ross Harrison's demonstration that explants of neural tissue from frog embryos, sealed into a chamber containing frog lymph, generated flbres which extended into the lymph clot. He thus showed, for the first time, thnt nerve axons are formed by the outgrowth ofprocesses from the neuronal cell body. These processes lengthen as a result of the amoeboid movement of thcir distal tip, now called the grgwth cone., This observation finally resolved one of the long-standing controversies in neurobiology, showing clearly that cuch nerve fibre is the outgrowth of a single nerve cell. Even earlier than this, mammalian cells (bull spermatozoa) had been frozen and stored for extended periods and recovered in a viable state for artificial lnsemination and selective breeding of cattle. This technology was to prove lmportant once the first continuous cell lines were.developed in the 1940s by

cclls can be released from 'suspended animation' on thawing and reculturing. As early as 1885, Roux had succeeded in maintaining embryonic chick cells alivc in a saline solution outside the body for short lengths of time. Although llris cannot be described as tissue culture, Roux's experiments opened the wuy for Harrison's pioneering work, where cells were not only kept alive, but

llurle and others, who were thus able to store cells indefinitely in liquid ttltrogen; at -L96'C there is imperceptible deterioration in cell viability and

glew as they would do in vivo. lly the early 1970s, methods were being developed for the growth of rpecific cell types in chemically defined media. In particular, Gordon Sato and lris colleagues published a series of papers on the specific requirements of dll'lbrent cell types for protein growth factors, attachment factors such as high rnolccular weight glycoproteins of the extracellular matrix, and hormones ruch as insulin or the insulin-like growth factors. The mixture of supplements rcquired for the serum-free culture of neural cells, defined by Bottenstein and Srto in 1979 (10), still forms the basis of many serum-free supplements for a widc variety of ccll types. Most of the additives in these supplements had

r
Coroline B. WigJey
otherwise been provided by ilr-defined and variabre biologicar fluid medium supplements such as serum, which is stilr a comp"""ir-.t'.any routine culture media today. over the next decade, tissue culture became, as well as an experimental research field in its own right, a technorogy useo uy marrf biotogists and biotechnologists in the newlyistabrished fieldif m"l;";iili;i;gy. Hybridoma technology alrowed the mass production of monoclonar antibodies from ceil culture medium; viral vaccine production depended on the mass culture
1: The cell culture loborotory
Trble 1. The early years of cell and tissue culture Lrto 19th century Methods established for the cryopreservation of semen for the aalcctive breeding of livestock for the farming industry tl07 Ross Harrison (1) published experiments showing frog embryo nerve fibre
growth in vitro lll2 Alexis Carrel (2) cultured connective tissue cells for extended periods and lhowed heart muscle tissue contractility over 2-3 months tlaS Katherine Sanford et al. l3l were the first to clone-from L-cells ll02 George GeV et al. (4) established HeLa from a cervical carcinoma-the first human cell line l164 Abercrombie and Heaysman (5) observed contact inhibition between f lbroblasts-the beginnings of quantitative cell culture experimentation ttl6 Harry Eagle (6) and, later, others developed defined media and described 6tt6chment factors and feeder layers ll0l Hayflick and Moorhead (7) described the finite lifespan of normal human diploid osllr It6:l Buonassisi et al. l8l published methods for maintaining differentiated cells (of lumour origin) llt8 David Yaffe (9) studied the differentiation of normal myoblasts in vitro

A recent development ig the cet curture field which points the way for future progress, has been tiie introduction of methods tor nisiotypic culture, using filter well inserts in culture vessels. This allows for the co-culture of two or more purified and defined populations of cells, ,"purut"Jurr=t in the same medium so that interactions -"diut"d by sorubre factors can occur. For example, the expression of differentiated claracteristi". noio"o'onstrated by a particular celr type in.isoration may be induced r, rri*rvpi. culture. It is hoped that such technology wilr furiher our understanding of the complex cellular interactions that oiiur within tissues in vivo,ro, orr] Juring develop_ ment, but also those which regulate tissue homeostasis and normal functioning throughout life. Many diseise processes are thought to invorve a perturba_ tion of cellular interactions (cancer is an obviour""*u-pt"j and the new methods now becoming availaLle for the study of cellular interactions in vitro will undoubtedly lead to progress in researctrin;; *ti"r, cause or contribute to disease processes. In the biomedicar field, theri are hopes that tissue culture technorogy wilr contribute more and more to clinicaf treatments involving transprants of various kinds. cultured skin ceils are already in use *o; or even as superior wound dressings in patients with ", burns and .kin".rtcers; curtured urothelium has been used to correct congenital defects in the urogenitar system; transplanted ceils carrying geneticaily engineered oNe ,"qu".r"", may be used in the future to correct Jefects in aiuuEti. puilrrt.,'puti"nts with cystic fibrosis or haemophilia, parkinsonian patients, #;-;;;some of the rare enzyme deficiency syndromes. There aie likely to be a ,r,.iob", of situa_ tions where it is preferable to correct a defective r.in"tion ,niu ffiunt"d cells rather than by conventionar.drug therapy. Even so, to, -ori situations, therapy will stlll involve administlring oiug, conventiona[y. In the case of complcx protcln hormone therapy] these-drugs wifl increasingry be pro'r duced from lurge'ncure ccil curtures, .ittrr or the appropriate ceri type or of

of host cells; recombinant DNA technology made use of curtured celrs, either as a source of or gene sequence or as an expression vector for recom'RNA binant DNA. In many areas ofrbiomedicar research, tissue curture methods were developed which replaced much of the former n""J io, unirnal experi_ mentation. For instance, in-vitro pharmacotoxicology became an established discipline for the investigation or drug activities of therapy. "*iir," J"rig" "f new forms

eclls cxpressing recombinant genes, in order to avoid some of the drawbacks

ol' microbial expression vectors. Biotechnology is still in its infancy and will rurrdoubtedly develop rapidly in the years running up to the centenary, in 2(107, of tissue culture's first appearance as a new methodology.

2, The laboratory
2.1 Design concepts and layout
't'issue culture laboratories differ from general purpose biomedical research

il;;;ii.,iuii"r""t.

Inlxrratories in that their most important function is to allow the sterile hnndling of cultured cells. It is practically impossible to work in a totally
gcrrn-free environment, so that provision must be made to ensure that cultures slrrl culture media are maintained free from contaminating micro-organisms. 'l'ltcsc will flourish in the cell culture environment and will rapidly overgrow,

kill, the animal cells by release of toxins or depletion of the rnrlrient medium. ln almost all modern laboratories, sterile handling of cell cultures is carried ortl in laminar flow cabinets or 'hoods', where only the operator's arms enter llrc stcrile work area. All other areas in the laboratory must be kept scrupukrusly clean, with surfaces kept clear of unnecessary clutter and swabbed tkrwn with an antiseptic cleansing solution at regular intervals. In addition to lhe sterile handling facility, all tissue culture laboratories must incorporate rurcus for a variety of other activities; these may be in one room or arranged as
arrtl in many cases

7
Coroline B. Wigley
a suite

7: The cell culture loborotory

bnch Glass-fronted cupboards for sterile glassware above

of course, budget.

of rooms according to the scale of the operation, available space and,

'dirty' operations, including disposal of used culture vessels and a washing-up facility for re-usable .qlalsware and equipment. In between, at increa-sing distance from the sterile handling area to produce a ,sterility gradient' within the laboratory, should be the other facilities essential to cet'i culture work. These are, in order

separation of the clean area for sterile handling of cells from the u.ei fo,

The sterile handling area itself should be positioned so that there is minimal movement of people past or through the clean area. There should be adequate
Fridge and Bench with cuPboards under for plastics etc. and mdia fridge

freezer

(u) (b)

Wash-uP and
Srrak

laminar flow cabinets incubators (CO2 and,

ttnks

(")
(d)

if required, dry,

ungassed)

preparation afea

Vllrier

storage space for sterile equipment and solutions instrument and equipment benches

lil[lllcrtion
ayrtErn

svf,t trhlk
lihrk and

(") (f)

Drying and sterilising ovens under bench

(g) (h) general

media fridges (sterile working solutions only) fud,ezer for culture reagents requiring storage at -20"C or below pteparation area for media and other solutions cold storage facility for chemicals and non-sterile reagents

(lttiner
Itlpotte wreher COZ incubators

balames, mgnetic stirreF pll mter etc'

double stacked

Where possible, in a separate room or rooms:

liquid nitrogen freezers for frozen cell stocks general preparation bench (k) storage area for unopened plasticware (l) sterilization oven and autoclave (m) drying oven (n) water purification system

(i) (j)

suitable for two or three people. llgure 1. Self-contained tissue culture laboratory,

(c)rrnelectroniccellcounterforlaboratorieshandlinglargenumbersofcell lines or for those involved in growth kinetics research (tl)airconditioning(thiswillbenecessaryfor'laboratoriesinwarmclimates e'g' laminar flow where many heat-generating items of equipment'
and Incubators may overheat cabinets, are concentrui"d irrto a small space.

if

(o)

washing-up area with sinks, soak-tanks, pipette washer, and automatic washing machine

theambienttemperatureistooclosetotheiroperatingtemperature) nitrogen (c) a cold room at 4'C for storage of media and for liquid
l'reezers

and sterilization facility. In addition to the essential facilities outlined above, several desirable features of a cell culture laboratory suite should be considered, according to the type and scale of the work to be carried out:

Plans of two possible arrangements are given in Figures and 2. The first is a single, self-contained tissue culture laboratory whilst the second is for a bigger group, with a clean culture laboratory and a separate, adjacent, washin!--up

(l)

cell lines and potentially some form of containment facility for incoming hazardous material receive biological

(a)

(b)

a warm room at 37"cfor large-scale cultures in e.g. roller bottles, timelapse photographic or video equipment, etc. a laminar flow cabinet away from the sterile handling area for the initial dissection of tissucs fur primary culture work

which A containment facility is advisable for laboratories 1i""., from elsewhere' Any potentially pathogenic ilnrrrples, including ir tttutcrialwillalsoneedtobehandledhereaccordingtoregulationsforthe comprise a sep-arate room' l,,,rr,rd they pose. A containment facility should containing a laminar flow without direct access to trr" r"st of the cuiture suite, a dedicated incubator which Iuxrd of the appropriui".ut"goty, e'g' class II' cells in quarantine-, c'n occommoOut" ."ufuti", gi.ruUt" iontainers to isolate (centrifuge, small to handle the cells ilncl flll back_up "q;;;;;;;eeded the laboratory, unless obtained directly ceh lines new to ,l"irig;roi.rr, etc.),
6

1: The cell culture loborotory

o

g
o G
3

6(d JO

E5

F3

FE

o ,t f

.o

g()ntaminants, should be isolated and handled as potential carriers

s8

co

bd E 5E
,aEs

(E

0)

from a reputable cell culture collection which guarantees them free from of mycoplasma, for instance, until shown to be clear. For further treatment of this whole topic, see Chapter 8.

o
.u

o)

t,2 Services
'l'lruue culture laboratories have some of their service requirements in comItton with more general laboratories, for example, water supply to sinks, gas lnpr lbr Bunsen burners, and above and below bench power sockets. Other Frvices are less widely used outside the culture laboratory.

ri t-o l8 to
lo lo tp
o
o o

o o
E
d

6 ! c (! .9

Ec
CL

o Oo_ 6=

o E d o o

o o o
a,

(E L

1,2.1 coz
Most incubators for cell culture work depend on a supply of CO2 to maintain n llxcd CO2 tension in the humidified air within the incubator (see below). lrlenlly, CO2 cylinders should be outside the main cell culture laboratory and lltc gas piped through, to maintain cleanliness in the sterile handling area. lllnce this may not always be possible, Figures I and 2 show cylinders alongrldc the incubators. A bank of CO2 cylinders should be secured to a rack ;tlaeed near the incubators and the gas fed via a reduction valve on the eylinder head, through pressure-tubing to the incubator intake port. ( irntrol of the CO2 level in gassed incubators is achieved through the Ittr:lusion of CO2 monitors, which increase the initial costs of an incubator but repny this with lower CO2 usage and better control and reliability. As long as llte incubator remains closed, usage is minimal and, in the event of a cylinder running out unexpectedly, the gas phase remains within tolerable limits for quile a long time. It is a useful precaution to have more than one cylinder on llrrc, with an automatic switch-over unit, operated by the fall in pressure as a lylirrder empties. 'l'hc effect of a leak in the CO2 system is worth consideration. If this is large :'rrorrgh and is left unchecked over a holiday period, for example, everything r'orrnected to that supply will be affected. Thus it may be worth having two torrrpletely separate supplies, each to a different bank of incubators. Valuable rloek cultures can then be duplicated in independently supplied incubators to svoid their complete loss in the event of a major leak in one of the CO2lines. lrrt'vcntion is always better than cure, however, and a little care in using the correct pressure-tubing and securing and testing all connections should avoid rrrort problems. A wash-bottle filled with soap solution (household washingrrp licluid is ideal), squirted around connections, makes a cheap and effective lenk rletector. In particular, newly connected cylinders should be tested in this wnyl dirt on either face of the connection can cause significant leakage.

CL

ED

sg OE
-st

=3 o
o I

.s o

Eb
-!E o.=

c c,
c
E
(U

o
OE
(E (E

o 3

o o

d; OE

o o

=O (l)o o(d
E

5> =o dg

EF
.:

E$ J6 38 o5 o6

q.
.n

': 3
= o
G

Effi., E
P

o
g f o 5 .2
al,

$$Hf Fe$gF#
o o

o
CD

s
P 5 oc
o

c*$eE
o

E c_

c o o

?g s(UI
_3

XE

E D
P

3ap

8E

-3

b8 6E 6

cDi (U-^

o=

9b (UE

tUo

o
:Q

cl
o o
E

th

-o

sb

oDo
Nr: tF llL

o g og ocL .=o 3(l,

o o

ssF

IL]

_8fi

o-l

2.2.2 tJltrapure water

Wutcr is used in the culture laboratory for several different purposes which tlsnand a particularly high level of purity. Water is used:

EE

Coroline B. Wigley

1: The cell culture Ioborotory

as a sorvent for curture media (when preparing from powdered formulae media

concentrates)

or

For a small laboratory with limited resources, it may be more cost_effective to buv readv-prepared, ri"gt"-.tt"ig;h'mediJ from a ,rt:til:, supplier. It wil be advisabre ats]o to uuy"rrttrup.rr" ,nui"i-iJriil o""urionat preparation of solutions of reagents which cannot be purchased reacy to use. In this situation' water from u Jonu"ntioiul double distillation apparatus may be adequate for glasswa^re rinsing urrO otn", general purposes. Ideally, a water purification slstem situut"o in the washing-up area can serve as a continuous lupnlv.or-ultrapure water for utt prrrpo."slirris witt atso help in the long-term tominimize the costs or media iro oti"i sierile, tissue culture reagents. Double glass distillation i, adequate ro, tt ni.i".tage in the purification of tap water but tras.oow-l-u-rgely,been " otTo:t (Ro)' This process typically ,"-ou", about superseded by reverse 9g% of water contaminants' Tap water fed to an Ro unit stroun nrst be passed ,rri."gr, water softener cartridge in hard water " ;;#;;. areas, to protect the Ro"rnventional The ourput from the- Ro unit u" ,tr"o (ror a rimiteJ time) in an intermediate tank which shourd"u.t b; ;;;; with hydrophobic srerilization filters on the air inlets. water from this tank can be used directly to feed the second-stage, urtrapurification system, which-is compriseo of a series ,f ;;;ilg"s for ionexchange and the remoya!_or organic Tr"iiir'"1;"" through carbon and other types of nrter. daier puri,y is monit#ed;rh resistivity meter which should reach about lg Mb/cm. IJltrapurification systems are " supplied by companies such as Millipore (Milli_e ,Vrr"_lo_ i.isons (Nano_ pure system)' Finally, ultrapure water is p)ssed tr,rougf, ;',ni"ropor" filter to exclude residual particulate matter, ln"i fulng micro_organisms. In the choice of water urtrapurihcation systems, the rate of pure generation is important, particularly if rarge volumes are needed water fQr a washing machine, for instance. vou Jrouti arso be sure that the unit wirl run water at hieh pressure and will not require an intermeaiate storage 'l'l'tap 'l'hc steririiation lircility' cycre rn."ro"u" .lm-iautomatic and easy to operate (ree bclow), rrherwise its impreme"i,i." *il auy, by buny lechniciuns and membra"" ;;;;;" may result. waror sh.urcr rro cotected uno uuto.iuuEl imm"oiatery for ster'e irrtcrnrcttlarc Hro*rgc t,nk can ,'L;;;;;ed to feed; il;;;;r" use. The washing nr'chlrrc [rr'rhe rrrrrrr,'trisritcd water'rinrl (see berow). Tanks and ,liglrt-tighr' Iuhlng llurultl bc fo prcvenr-"ig"l "y.r"* growth. Alth,ugh the c'rir ,r' tirc Ro unit i,, o ,,"uu.tuntiar initiar outray, provided lltrt c're is r'ken r'ereurrse rrrtr slcririze the unit and its ,"rt riin'" at regurar

for supplements to culture media o for the final rinsing steps in washing up re-usable glassware which will used for the preparation, storage, a"ndianOting of culture media

as a solvent

be

inlcrvals e.g. monthly, with a proprietary solution containing formaldehyde, thc nrcmbrane will have a lifespan of several years. There is an automated lerrrrsing cycle on most models e.g. Fisons Fistreem RO 60, which requires rRlttirnum effort and cost to run. The initial cost is soon repaid in much fetlrrccd power consumption, in the saving on replacement of elements etc., gntl in time spent on cleaning and descaling a glass still.

il ilil;ents

3, Equipping the laboratory g,l Maior items of equipment and instrumentation

9,1.1 Laminar flow hoods It wus common practice in the early days of tissue culture to carry out sterile ploccdures under a sheet of glass or Perspex, clamped horizontally about
in a still-air room. Bench and glass sheet were swabbed with 70o/o ethanol and, since most microbial contamination falls from drrwrr gltovc or is wafted into an open, sterile vessel from a non-sterile source, Gf,lclul manipulation of sterile material under the plate glass was often sucelslul. Nowadays, however, most tissue culture laboratories rely on the use uf lrrminar flow hoods for all sterile handling work. The alternative is to rupply whole rooms or a suite of individual cubicles with sterile-filtered air. T'ltc rooms are kept under slight positive pressure to ensure that non-sterile alt' is not drawn in. This arrangement may be necessary for some large-scale ptoccdures where cost is not a major consideration. t,rrrninar flow hoods, then, are widely used in cell culture laboratories. Tltcy operate by filter-sterilizing the air taken in, so excluding particles lltcluding bacterial and fungal organisms. Air thus sterilized then passes in a verlical downflow on to the work surface. Some other types of cabinet pt'rltluce a laminar flow of sterile air which is directed horizontally towards the .tl)clator, but these do not offer protection to the operator from the potential Irnzrrrds of some cultured cells. l,irminar flow cabinets are classified according to the degree of safety they ol'l'cl the user and it will depend on the work to be undertaken, which is most appropriate (see Chapter 2). The principles behind the design of most laminar flrrw cabinets are shown in Figures 3 and 4. Figure 3 represents a standard lattrinar downflow cabinet, and Figure 4 shows a cabinet with Class II safety rlrrrrtlards, suitable for handling primate or human material, some virally litlcctcd cells, certain carcinogenic reagents, etc. Both types of cabinet provitlc similar control of air-borne microbial contamination. In addition, the ('lrrss II cabinet offers greater protection to the operator by:
0,,1 ttt above a bench

;dT;;**

b".b;;;;iiilotn".

(rr) provision

of a front window

(screen) with a lower edge which gives

minimum turbulence to air drawn in from outside, whilst allowing adequate flccess for the operator's arms. The air taken in at the front acts as a safety curtain, prevcnting acrosols of potentially hazardous material from

Coroline B. WigJey

7: The cell culture laborotory

Air exhaust through outlet filter

Fan

Main H.E.P.A (high efficienCy

particulate

air)

filter

$/
Perforated grill

Laminar downflow of sterile air

Air /I exhaust

,rrl

/\
/

/r
wo.

Laminar downflow ol sterile air

rrrt"".

\
-t----------> llgure 4. Class ll laminar flow cabinet, for sterile work requiring higher operator safety
Stanrlurds than provided by an open-fronted sterile work cabinet lsee Figure 3l..

\-----=--* [
Flgurc 3. Laminar downflow sterile work cabinet.

(b)

reaehing the operator. In an empty hood, this curtain of air is drawn immedlately to exhaust, sodoes noi .'o-pro-ise the steriliffi any material innicle tlrc working area. However, the operator,s arms disturb the air flow and eaune eddies. N.on-sterire air may then enter the cabinet and it is wise to avoiel uairrg the fr,nt r0 cm of work surface for sterile rrananng. exhauetlng air lrowing ncross the curtures to the raboratory only after pessing. it througrr sn uutrer firter; most i, ,";y;i;l cauinet as nhown in the dlcgram

{c)

rrn air flow monitor and alarm system which warns when the rate falls bclow or rises above safe levels

;;;;ii'e

'l'ht:se and other types of cabinet are discussed further in Chapter 2. All cabinets should be inspected regularly according to current regulations gnrl rnnintained through good laboratory practice; in particular, all spillages rltorrltl be cleared up immediately and the surfaces wiped down with 70% ellrirrrol (it is useful to keep a wash-bottle of this close at hand). The cabinet :trurrld also be cleaned out completely at regular intervals, by removing the elelneltuble work surface.

10

l7

Coroline B. Wigley

1; The cell culture laboratory

when choosing a laminar flow cabinet (or hood), a number of factors should be taken into consideration. The first, of course, is that the hood should be of the appropriate class for the work which will be carried out in it (for further discussion, see chapter 2). For a general purpose tissue culture laboratory, a class II hood is generally the most upprop.iute, providing a degree of protection to both worker and cultures, so plrmitting tire handling of all but the most hazardous cell lines. Such hoods aie availab-ie in a variet| of sizes (usually based on nominal width) and can vent either to the laboratory or can be ducted to the outside. Ducting to the outside need only be considered if there is good reason to do so, e.g. regular use of higher risk cell lines or the frequent need to fumigate the hood. Most tissue labor"iltur" atories will find a 4 feet (120 cm) wide hood adequate for most purposes, although certain procedures, particularry at large- oipilot-scal", -uy be more easily carried out in one that is 6 feet (1g0 cm) wide. It should be noted that the working height within such hoods varies between manufacturers (an important thing to consider if frequently dealing with large bottles or spinner flasks) and the width is only a very rough guidi to the riseful working area. The actual internal dimensions of two class II hoods from different manufacturers, both nominally 720 cm wide, are given below.
Manufacturer

I I I

lslllPcrature
gtu phase hurrridity

Width (cm) Depth (cm) Height (cm)

Manufacturer B
105.5

720.5

59.5
82.5

50.0 68.5

facturer B. Other factors which may be considered are listed below.

clearly, the hood from manufacturer A is much the larger, with a working area 36o/" greater and a working volume 640/o greater than that from manu-

(a)

Does the hood require an internal power point and/or gas tap? (These can be fitted on request by most manufacturirs if not s.rppli"d^u. standard.)

(b) Is the inside of the hood easy to clean? There should be no difficult, inaccessible spots and the work surface should be easy to remove and
replace.

(c) will it (tl)

nlrrnrh, 11.1.2

be necessary to open the front of the hood regularly to introduce large vessels? If so, the front glass, once opened, stroua remain open by itsell'so that both of the worker's hands are free to handle the vessel. woulcl it bc useful for the height of the working surface of the hood to be rudjustnhle? lf so, some manufacturers can supply height-adjustable

Incubulora

llrcubatorn lirr eell cullurc work shuuld provide an,environment in which the lirllowing nro contrullod:

The tcmperature required for most mammalian cells to grow optimally is 37"(', lncubators should be set half a degree below this for safety and a cutEBt ogrcrated at 39oC, above which cells rapidly die. Avian cells prefer a dlghtly higher temperature, reflecting the higher body temperature of birds, *hlh,. rnammalian skin cells, for instance, prefer a slightly lower temperature. Eellr liom cold-blooded animals are generally incubated at a constant tm;trrature within their normal exposure range. Mrlnl cell culture work requires that cells are incubated in an atmosphere *hlclr contains an elevated CO2 tension. Culture medium is thus maintained at t phyriologicalpH (7 .2-7.4) by the equilibrium established between dissolved O1 nrrd the bicarbonate buffer in the medium (see Chapter 3). The gas phase n hc ldjusted in culture flasks by introducing the correct gas mixture from E eylirrrlcr via a plugged pipette and then closing the cap tightly. Alternatively, Eptt vcssels can be incubated in a sealable chamber, which can be humidified, ga$crl, und closed. Several suppliers market such chambers e.g. ICN BiomediBlr, lrut a good workshop should be able to produce an equally effective vcrrhrrr l'rom Perspex, with a removable front fitted with a silicone gasket and bUtterlly screws to rnake a tight seal when assembled. Inlet and outlet ports Wltlr tubing that can be clamped permit the introduction of a gas mixture. Such hantbcrs then allow the use of less expensive, dry-air incubators. Monl tissue culture laboratories, however, use CO2 incubators as the most eeGuullc and reliable wayof providing the right incubation conditionsfor cells in Vtttctl vcssels. Indeed, for certain critical, sensitive work, a CO2 incubator may be eencntial. Even in tightly capped flasks, gas-phase leaks can occur. Whilst in Bthet' irrcubators this can lead to a (sometimes devastating) rise in pH of the medirrrn due to loss of CO2, this is not a problem in a CO2 incubator. ('( )., incubators can regulate the CO2 tension in a number of ways. In olderrlylc irrcubators, a constant flow of a CO2lair mixture was utilized, with each gar lreirrg supplied separately from a diffrent cylinder and mixed according to rrl proportions by gas burettes. The major drawbacks to this arrangement $ert' lhc cxtravagant use of CO2 and the likelihood that, if the CO, ran out unexpcctcdly, the air supply on its own would rapidly flush out the incubator arrrl rrllow the pH of media to rise. 'l'lrc rrtust efficient and widely used method involves the use of an instrumettl wlrich measures the CO2 level in the incubator and activates a valve tu rltrrw on pure CO2 from a cylinder, whenever the gas phase falls below a ael vuluc.'l'his is generally 5%, although media containing high levels of bh,allrolrute, such as Dulbecco's modified Eagle's medium (DMEM), particulet'ly wlren used for serum-free cultures, should be incubated in a lOY" CO, Btntorltlture (see Chapter 3). Simple CO2 calibration devices, e.g. the
13

t2

Coroline B. Wigley

1: The cell culture loborotory exposed to Wlth a timer for long runs with sensitive cells which have been a benchtop microfuge is a useful laboratories, ffJteotytic .nry-"r.hor some which lgai,iun for high-speed centrifugation of small volumes of reagents thawing from the freezer' my gcnerate Jprecipitate' e.g. after

Fyrite test kit marketed by Shawcity Ltd, can be purchased to check the monitor readings, but for most purposes, the colour of medium in a dish without cells is an accurate guide to the experienced worker. Humidity is maintained by including a water tray in the bottom of the
incubator over which the air is circulated (most incubators have fans to ensure even temperature and humidity throughout). Water trays should be prevented from harbouring micro-organisms by the inclusion of a non-volatile cytostatic reagent such as thimerosal (but not azide, which reacts with metals) or a low concentration (1%) of a disinfectant detergent such as Roccal, supplied by Sterling Medicare. The water level should be topped up regularly (using deionized or RO water) and condensation aspirated from other areas of the incubator base. There is, nevertheless, a potential risk offungal growth in particular, in humidified incubators. The interior should therefore be cleaned out and wiped over with antiseptic at regular intervals; it is important to check when buying a new incubator that the interior can be dismantled to allow adequate cleaning. LEEC Ltd make simple, reliable, and easily cleaned CO2 incubators at the lower end of the price range. When choosing an incubator, it is nevertheless worth considering whether the extra expense of a model with an automatic sterilization cycle (e.g. Heraeus Cytoperm), is justified. This may be a distinct advantage if the tissue culture laboratory has a high number of trainees or where there are other factors which make contamination more likely. If the whole culture laboratory is to be fumigated periodically, care should also be taken that the CO2 monitor can be removed from the incubator or is of a type that will not be damaged by aldehyde fumigants. For special applications, CO2 incubators are available in which the partial pressure of oxygen can also be controlled. As well as CO2, such incubators require an oxygen and nitrogen supply. There are also CO2 incubators that
can be used in high ambient temperature conditions (e.g. laboratories without

9,1,4 Microscope

quality inverted An esscntial instrument for any culture laboratory is a good optics and a photographic facility' mia,.,r.op", preferably with phase-contrast membrane blebbing, b-e!l nlorittoiogy-ttte degrei of spreading, granularity, and so on-should be monit-G pr,rpfrtion"tf multinricleates,-vacuolation, ior*.f ,"g"furly for signs of stress in cells. Cell morphology is a sensitive lndh,lrtor of problems with culture conditions' good Eilrty signs of microbial contamination can also be detected with a Regular checking of cultures under the microphArc contrast microscope. by allowleup. .un help to avoid catast.optric losses of irreplaceable material at in early stage. Also, cells which are used for ing'n proUte- to b" noticed tfperitn"nts in a less than healthy state may give variable or erroneous working ieiutr*. When choosing a microscope, select the long or extra-long flasks and even roller bottles can be viewed' There dktnlrcc condenser so i-hat too ir urur'lly no need for objectives above 20x; their depth of field is often of all but the very flattest cells. A_ good, lowIttw 16 ,ibtui.r a sharp image very useful for scanning fu**r, wide-field oblective, e.g. the Nikon 4x, is for instance' Lulttrrcs for foci or colonies,

8,1,6 l'ridges and freezers

air-conditioning or that are constantly exposed to the sun or other heat sources). These use heating or cooling, as necessary, to maintain their set incubation temperature.
3.1..3 Centrifuge The main use of a centrifuge in a tissue culture laboratory is to spin down cells during tissue disaggregation or in harvesting cell lines for subculture, freezing, analysis, etc. A very simple benchtop centrifuge which will reach 80lfi) g, preferably with a variable braking system, is generally all that is needed. Various considerations may indicate the need for a more sophisticated piece of equipment. The volumes of cell suspensions to be handled in a single opcration will dictate the bucket size and adaptability needed, for instancc the capucity to take four swing-out buckets each holding five 50 ml tubes in onc run, tbllowed by n mixture of 15 ml and 50 ml tubes. If primary culturcg urc antlcipnted, il ntuy bc an advantage to have u ctmlcd centrifuge
74

no ice box) are FEI rnost laboratories, domestic larder-type fridges (with requires considerable space and i1 -uy be more ederguate. Media storage ggxyl:nient to store ,rnop"n"d bottles in a nearby cold room, if available' The media in irldgc in the tissue culiure laboratory can then be reserved for opened bottle designated for one individual's work U,Uricrrt use, with each culture wltlr orrc cell line. Ideally, separate fridges should be used for sterile solutions, chemical stocks, etc' lnerlirr lnd for non-sterile of sera, lircczers at -20"C and, ideally, -70"C, will be needed for storage reagents which ire unstable at higher temperatures' At the lriluli0ns, and which is prone to lowt r, tcmperature, sera and proteins such as collagenase, periods. Some of tlegrrrtlation even at -20oC, can be stored for extended co2 or liquid llreEc lower temperature freezers are available with liquid facilities that permit temperature to, be maintained even in iitt,,,1,"n back-up nitrogen freezers lhe rvc,nt of an eiectrical supply or compressor failure. Liquid cryopreservition of cells will be considered in Chapter 4. f,r lrrrg-term

S,l.ll Miscellaneous small items of equipment

oC, will be needed to warm media etc. before use, to A wltcr bath, set at 36.5 incubations. thaw l'rozen aliquots of reagents or cells, and to carry out enzyme

16

Coroline B. Wigley
Care should be taken to change the water regularly (e.g. at weekly intervals) and to add a cytostatic reagent such as thimerosal (but not azide) to prevent microbial growth" A submersible magnetic stirrer is a useful accessory. Some workers prefer not to use water baths because of a perceived risk of contamination from the water and associated aerosols. In this case a small dry incubator at the same temperature fulfils a similar function but has the disadvantage that heat dispersal is less efficient. In the area for preparation of reagents for culture use, there will need to be a balance, a pH meter and one or more magnetic stirrers. The balance should be capable of weighing accurately in the milligram range. Many reagents for culture work are active in the microgram or nanogram range and stock solutions are usually prepared at 100-1000x. An osmometer is a useful, but

1: The cell culture loborotorY

llulion, kept in a wash-bottle standing on a plastic surface (since- it corrodes m*tut;, rtro"to be drawn through the aspirator line at the end of each work to avoid both reaci0rr and between handling different cell lines. This will help contamination attd cross-contamination between cell lines via the Fierobial lfpln,tor titt". The aspirator jarcontentsshouldbedisinfectedbeforeemptying, separate room. As an alternative to an aspirator tret'eraUty into a sink in a u**"-'Uty, if there is a suitable water supply at sufficient pressure nearby, iar L-':lnrpte tap syphon is a cheaper and more convenient way of aspirating fpent'rnediu-,- iinc" waste is drawn directly into the mains drainage.

1,3 Culture plasticware and associated small

consumable items 8,8.1 Tissue culture Plasticware Thlr is supplied by a number of specialist
companies; the. top four or five (usually polyCenrp,,ni"r produce very comparable ranges of sterile plastics different modificaiiyrl',"y, in terms of boih price and quality, but may offer commonly iienr.ttto basic design or iniroduce new, specialist ranges. The most Eted ilcms of Plasticware are: to produce an electroia) lirr growing cells (most are 'tissue culture' treated,adhesion): clarged'surface for wettability and cell It.tically

not essential, piece of equipment, especially if media are prepared from powder or basic ingredients or if a number of additives are used which will affect osmolality. The Fiske One-Ten osmometer supplied by Astell Scientific, for example, measures small volume samples (10 pl) and is quick
and easy to use. Spent medium can be drawn off into an aspirator jar, most conveniently situated underneath the laminar flow hood, with a pump to draw a vacuum. Figure 5 shows such an assembly, which can be adapted to serve two or three laminar flow cabinets by using a two- or three-way connection on the aspirator line and an in-line valve to close off lines not in use. A simple diaphragm pump will pull a vacuum sufficient for one aspirator line but a more powerful pump, such as an oil vane vacuum pump, may be needed if two or more lines are open simultaneously. The aspirator jar should contain a small amount of detergent such as Decon

llat-bottomed flasks

o Pctri dishes o ntulti-well dishes

(Decon Laboratories Ltd), and a small volume of Chloros (hypochlorite)

r conical flasks (for suspensions) r rollcr bottles r lubes

{b) tirr handling solutions and cell
suspensions:

Disposable Pasleur pipette

r volumetric PiPettes r plastic Pasteur PiPettes r rrricropipette tips r ccntrifuge tubes

storage:

(e) lor

r r
(

sttmple tubes

lippendorf tubes o cryotubes o hrgcr volume screw-capped bottles

Flgutr

l,

Arpltator lrr mrrmbly for withdrewing spent medlum. 10

)ttr ttl'the most useful vessels for growing up stocks of cells in culture is the 25, 75, and \75 cmz . flgrh , which is widely available with uiable sittace areas of necks and are supplied with a convenFlgrkr huve cither itraight or canted incubations' tlprt,,t. two-position, sciew-threaded cap for vented or closed
77

Coroline B. Wigley

7: The cell culture laborutory

ccllulur,matrix protein substrite. consideration of such ,p""ih" ,rurtrates is tlertll with in chapters.4 and 6.It may be sufficient to coat the tissue culture *rrliree a highry charged porymei such as porylysine o, potyornithine _with at rulr'ul l0 pg/nrl. 'l'his is also useful for encouraging attachment oicem wnich ttttuully grerw in suspe nsi.n, for e .g. Hoechst staining for mycoplasma testing.

For some cells, the tissue curture surface, even when treated to enhance cell adhesion, does not provide an adequately adhesive surface. some ceils have an.absolute .requirement, especialfu in serum-free medium, for an extra-

area. cells grown in suspension b" curtured either in rpinn", nurrc, where the "un medium is constantly circurated by a magnetic stirrer, in flasks on a rotating platform (kept in a warm room Jr dedicated "o.ri"ui in""ou-t-o4, or in static culture on non-adherent plasticware. There are a number of other new developments in the design of curture vessels for particula, ,r."r, such as chamber slides (e.g. Labtek, from Nunc) for growing ce's to Le used for immunocytochemistry, or tripre flasks (Nunc;, "wtricrr ;;ee stacked surface areas for cell growth within a conventionally ;;;" sized n"*t, a solution to the problem of limited incubator space. Supplieis ", of tissrre crrtture plastics will advise on new or more specialist items on ih"i, li.ts and most catalogues give detailed information on the specification and uses of their pioducts.

applications) and 24-well are the most widely used. Four- and six-well plates may have specialist applications, such as the accommodation of filter well inserts (e.g. those supplied by costar) which alow oin"r"niiypis of cells to be.grown^separately, on one or more membrane ,upport, *iitin the same volume of culture medium. For large-scale adherent cell cultures,.prastic roller bottres can be used, which require a special apparatus within a cabinet or warm room to turn the bottles at a constant rate, thus utilizing a large surface area for growth in a relatively small volume of mediu-. in"r"uringly, this typ" large-scale culture is achieved in other ways, such as the us=e "r bt,oi".o"urrier ueads (e.g. 'or those supplied by pharmacia Sera-Lab;, o,utri""r-"in*r"g'r"growth of cells, or
meshes of various types to increase the available culture surface

Alternatively' some suppriers (e.g. costar, Nunc, and Greiner) may offer flasks with a sterile, hydrophobic firter in a perforated cap so that gas exchange can occur without Lven the slight risk of microbial contamination associated with a loosened cap. Tissue culture dishes are widely used and are available in 35, 60, 90, 100 and 145 mm diameters with a viriety of modifications possibll, such as a series of internal wells and.grids. Genlraily, the lids are vented for adequate gas exchange but designed for minimum Lvaporation. Multiwell dishes are supplied in various sizes, but 96-well (flat or round bottomed for different

llltcring small volumes of tissue culture solutions. The pore size should be
0,2 pm or even 0.1 pm in order to be sure of excluding mycoplasma as well as

other micro-organigms. For most purposes however, 0.2 pm is the standard pore size for liquid sterilization. It may be necessary to use an assembly with a prelilter of e.g. 0.8 pm, if solutions are likely to block a low pore size filter tsrdily, e.g. a semi-purified enzyme preparation or a biological fluid.

Scveral features should be considered before choosing the appropriate flltcr. Some filter membranes e.g. those made of polysulphone, are low protcin binding and essential for proteins where the concentration of the
f,llcred solution is critical, particularly if the molecule is highly charged, as are ilrmc of the polypeptide growth factors. Very small volumes of valuable rotqlonts should be filtered through units where the 'dead' space (hold-up Volume) is minimal. It is cost-effective, therefore, to have some of these more rpcnsive filters reserved for special purposes. Larger volumes can be filter llorilized using either pre-assembled, disposable units which attach to a vncuum line (see Section 3.1.6) or, more economically, a washable, autoelnvuble unit in which the filter can be replaced. l,nrge tissue culture facilities may consider production and sterilization of ftotlia 'in house', in which case a stainless steel pressure vessel, connected to a nlllogcn line, may be useful. However, this is outside the requirements of motl tissue culture laboratories. Filter sterilization is considered in further tlslnil in Chapter 2.

E,l.ll
lll

Pipettes

'l'hosc needed for all culture laboratories are of four basic types, with differttscs:

I I t I

ttticropipettes with disposable, autoclavable tips trrrplugged Pasteur pipettes

plrrgged Pasteur pipettes

volrunetric pipettes

Plnntic micropipette tips are supplied by a number of companies to fit the vttt'lt'ty of micropipettes on the market (see below). These will be of several lyltcs, tn suit volumes in the 1-20, 20-200, and 100-1000 pl ranges and up to 5 trrl. ln addition, extended, fine tips are available for use with small volumes lil rrru row diameter tubes. Tips also come with or without a circumferential
t,lrlp,r'

part way along their length, so that they can be supported in an

3.2.2 l'lltcrr ltlr rtorilizing tissue culture solutions sovcrsl ,rr,rppllc*r r|. nre,ririzing filters (Miilipore, German, etc.) produce dispunrble, eterllo urrlltl I'o' nttnchment to a syringe, which are'suitable for
18

urrlocluvable tip carrier and easily attached to the micropipette without handllttp, Scvcral companies also market tips with an integral filter for extra safety Itr xlt'rilc handling, but these may have to be purchased pre-sterilized, which urhhi lo their cost. Autoclaving, but not irradiation (which is in any case not I'erurible lbr most laboratories), causes the filters to deteriorate. ml, with an integral bulb) and l)ir;rosuble plastic Pasteur pipettes

(l

1g

Coroline B. Wigliy volumetric pipettes of all volumes (1-50 ml) can be purchased from m tissue culture plasticware suppliers in pre-sterilized packs. However, this is avoidable expense if substantial numbers of pipettes are used and there i facilities to wash and sterilize glass pipettes. Generally, glass pasteur pipet
plugged and unplugged glass Pasteur pipettes will be needed and are intended for disposal after a single use. Extra-long unplugged pasteur pipettes are preferable for reaching to the corners of large flasks for aspiration of spent medium. Standard length, plugged pasteur pipettes ur" u."d for any small. scale, non-volumetric pipetting of tissue culture solutions and for the introduction of gasses into e.g. a flask or sealable chamber, where the cotton

7: The cell culture loborotory

ggg, but this is not usually a problem in practice, provided that high quality brolilicate glass is used and that washing-up facilities are good. Glassware hEE a g.eutet propensity to adsorb substances such as alkaline detergent on to

are supplied non-sterile but ready to use after dry

heit sterilization. Both

It! turl'ace, than plastics such as polypropylene. Whatever the choice, most

plug ensures sterility. Glass volumetric pipettes should be of good quality borosilicate glass, graduated to the tip, preferably with the maximum volume graduation at the top. A relatively wide-bored tip is advisable for fast delivef. A selection of volumes from 5 to 25 ml is generally required; below this, micropipettes are more accurate and convenient. The most frequently used pipettJs are 5 and L0 ml volumes and there should be at least nve timei the num-ber in daily use, to allow for those out of circulation in soak, wash, or sterilization. Re-usable volumetric pipettes will have to be unplugged for washing and replugged before sterilization, care being taken to discard any thaiare cracked or broken (see below).

prcparation and bottles for storage of sterile solutions. Schott bottles fBehott Glass Ltd) are particularly suited to this. They are robust and will Itand the pressure created by solutions autoclaved with a tightly closed cap (,g, hicarbonate). The cap is deep and will therefore ensure a good depth of Ittlle outer screw thread surface and greater safety on the few occasions then pouring cannot be avoided. 911,0 Miscellaneous small items I plpctte cans

tebcrutories will require

selection of volumetric cylinders, flasks, and beakers

I ! I I I I I I !

luhc racks
heenrocytometers

lntlrument tins plpcttc bulbs

Fgr tlrc preparation area:

Buloclave tape

For those on a tight budget, conventional rnanual pipette bulbs will

Depending on the number of regular users, sets -uy n"id to te replicated. In addition, it is useful to have a multichannel micropipette; this is essential if much work is done with 96-well plates. For many liboratories, an automated aid to pipetting larger volumes is also a high prioiity. several companies make, lightweight, hand-held, battery- or mains-operated units for this purpose, with loading and dispensing push-button controls and an air filter toriteritity.

Formostlaboratories,arangeofmicropipettescoveringe.g. t-zop,i,zo-200p,1, 100-1000 pl, and 1-5 ml will be adequate, perhaps *ith orr"'o, two fixed volume micropipettes, such as a 50 or 100 pi for itequently used volumes.

and needs. Some are autoclavable for additional safety for sterile procedures.

3.2.4 Pipetting aids A number of micropipettes are available on the market to suit most

evcn tape Erowne sterilizer control tubes (for oven and autoclave, from Solmedia) Cotlon 'rope' for pipette plugging
Bttltrclave bags (at least two sizes)

lf rc-usable volumetric pipettes are used, it is convenient to sterilize them in clttinless steel, cylindrical cans or aluminium cans with a square crossltion, according to preference. Aluminium cans are less expensive, but will Eildizc over a period of time. There should be at least one can per worker for EBelt pipette volume-enough so that individual workers can keep a can, once
gperrcd, for their exclusive use. Shorter cans are needed for Pasteur pipettes.

be adequate; these are available in a range of volumes. una-er no circumstances should mouth pipetting be considered for tissue culture work.

Elverr-stcrilizable, tightJidded flat tins are convenient for sterilizing instru. ments for dissection, sterile handling of coverslips, etc.

containers are lers prone to lenching trace elements which might be deleteri20

3.2.5 Glaegwaro for tissue culture use Laborotories diffcr in their preferences for glass or plastic for handling and storage of solutions for tissue curture use.' There ii no doubt that p'iastic

4. Washing re-usable tissue culture equipment

tissue culture equipment which can be washed, sterilized, and recycled rhorrlcl go through the following general process, either manually or through a tlrurrc culture-dedicated automatic washing machine with both acid-rinse and

All

dlalilled water-rinse facilities.

2t

Caroline B. Wigley

1: The cell culture loboratory

Protocol 1. washing and sterilization of re-usable labware

Eguipment und ,e"gents

Ultrapure water

o Chloros (hypochlorite) solution for soaking r Detergent [e.g. Micro (lnternational producti Corp.) for manual washing; low_foam foi
mach i ne, as recom m ended bl m a nufi"tr."ii

o AnalaR. (or similar high purity) HCI for manual washing; formic or aceiic acid for
machines

e Rigid plastic soak tanks, cylinders, and

oeakers (tor small items and instrumentsl

f,odical flats, bijou bottles, etc. and one for fine glassware such as beakers Entl cylinders which are easily broken. The soak tanks can be filled with tap tvater but must contain Chloros or an equivalent sterilizing solution. This ihould be changed regularly. Presept tablets (Johnson & Johnson Ltd) proVlde a convenient way of making up this sterilizing solution. If the washing-up llgu is outside the main laboratory, it is acceptable to collect glassware in, for Iutnple, a bucket kept beside the work station, provided that it is taken to tha soak tank as soon as possible at the end of a session. Glassware for
EBlking should:

Method

1. Soak items either in Chloros solution (except metals), or directly in detergent.

2. Wash with detergent (by soaking or machine). 3. Perform tap water rinses (continuous or seguential). 4. Rinse with acid (except for metals). 5. Repeat tap water rinses as step 3 above. 6. Rinse with distilled/reverse osmosis/ultrapure water (two or more times).

I I ! I

he rinsed under a cold tap hnvc any tape removed ltnvc marker pen labelling removed with acetone (a carefully labelled washItottle kept by the tank is useful) lrc immersed completely in the soak tank

l,l

Washing
e.g.

lVhen the tanks contain a load for washing, glassware can be taken straight

7. Dry with hot air. 8. Store temporarily, capped or covered, e.g. on preparation 9. Prepare for sterilization (see Chapter 2).

fttiln the tank. Different laboratory washing machine manufacturers,

bench,

ll.

10. Sterilize by autoclaving or in dry oven. Store for use (in dedicated cupboards or racks).

Problems with inadequateJy y?:h"g equipment are common, avoidable by attention to the following.'

but mostly

4.1 Soaking
Many tissue culture sorutions.contain protein. If glassware is left unrinsed and unwashed for any length of time aftei use o, i, illo*"J ;;;;;r, there are several consequences:

Lgttcer, recommend different detergents, usually liquid and often their own brnlrtl, and acids for acid-rinsing, e.g. 507o acetic acid, which is further dlluteC in the machine. When loading the machine, care should be taken to flrure that all items, especially caps and small bottles, are stable in the 'open lldc down' position and can be rinsed adequately. lf gl:rssware (or dissection instruments, for instance) are to be washed by hntttl, they should first be soaked overnight in a detergent solution. It does Fol nrrrttcr, in this case, if the detergent is not low-foam, but it should rinse off ontplctcly under running tap water. Some products e.g. Micro (International Ffltrlucts Corporation) are advertized for their suitability for culture work or pfee ision equipment use, particularly where protein-contaminated material is lllely, ll'possible, for non-metallic equipment, an acid-rinse step should be

r
a

residual medium will support microbial growth and act as a source of airborne contaminants in t-he laborutory-- -

a driccl<ln protein

wiil be difficult to remove by normal washing procedures n sterlize d, *il n ating il I 1xili, l:l1,,;,vi11: l"yT." : _*_h"" r,"i_oenatured j:li:,tj:0""1 or proteins, llli:]lll ll1-::il c$rbonlt) finto new solutions ",:llfrH:

;";;;;mi

inclutlccl to neutralize residual alkalinity (although some detergents are Ehlrlccl to have a neutral pH so should not require acid-rinsing). Further tap *[l* rinsing should be followed by two to three changes of double-distilled, R() or', better still, ultrapure water. (llnssware must then be dried in a hot air oven, set at about 100"C (some ilnnlrirrg machines incorporate this facility), before re-sterilization. Clean llettts which cannot be sterilized immediately must be kept covered or inVelle(l on a clean surface until they can be prepared for sterilization.

pnrccdure$ ure therefore advisable. Soak tanks should be situIted near lhc wurhlng_up arear one for heavy-duty glussware

l'hc folkrwing

{,il Washing pipettes

Re.rraublc glass volumetric pipettes can be soaked in a plastic cylinder conte lrrilrg ( )hloros solution, kept beside each work station. Suppliers of laboratory
2S

such

as

22

Coroline B. Wigley
plastics, e.g. Nalgene, from Nalge Company, market all the items required for the procedures described below. In principle, the steps involved are those outlined for general glassware above, but several points should be noted which are special to pipette-washing. After a load has accumulated in the soak cylinder, the protocol below should be used, unless a pipette-washing facility is available in the automatic washing machine, when supplier's instructions should be followed,

1: The cell culture loborotorY

!,

General care and maintenance of the tissue culture laboratorY

shourd be made.here'

{ft t-1 "t]:-t"^|: ^t**::: rhe iil;,;;;;;;;;i;t# ?TryI.:i: :::3::t: an d cre anrin" * ; : il ;#";;i"" ;"';h;';;;;' tissue cultu'" "i :l-" 1i!:.1"t".'{::."" :111 "f iilffifi;I" "'p**uv'o *ti"r, must be completed on a tl"ll,"::::L':: lrrrl'r'rtcrtr' "-'*t-"il'u.r., i1 .u. l?!:it::v: daily, weekly,
and mainten.ance Bontc aspects of laboratory care Opentte a check-list ol monthly, etc., basis'

Protocol 2. Washing glass volumetric pipettes

1. Unplug the pipettes using forceps or a high pressure air or water jet

""" tnd cxtended to suit the individual laboratory'

Enlly checkJist:
'

'i1iil"i,i,"iil; ril;

modified be used as a guide but can, of course, be

2. 3. lnsert the carrier into an outer cylinder containing detergent solution

and leave overnight to soak.

applied to the opposite (tip) end. Place them, tip upwards, in a plastic pipette carrier.

4. Transfer the carrier to an automatic, water siphon-operated pipette

washer."

I I I I I

fcc.rd incubator

temPerature'

clteck CO2 cYlinder Pressures ttllt uP water baths

5. After 3-4 h, add about 50 ml of concentrated acid (HCl or acetic) at the filling stage of the cycle and turn off the tap when full, 6. Leave for about 30 min to remove residual traces of detergent. 7. Run the tap water rinse cycle again for at least t h. 8. lmmerse the pipettes in their carrier in two or three changes of ultrapure water, in a clean plastic cylinder kept for the purpose.

a etnpty PiPette soak cYlinders t elteck glassware soak tanks for washload
etc' rrplcnish stocks of sterile glassware, solutions' further details and additional items Thln is Only a brief list of essential points; 4'1' Bfe covcred in Chapter 4, Section
Wpel,t.v check-list:

tlttw Chloros through aspirator lines

9.

Drain for a few minutes and transfer the carrier to a drying oven.

' Pipette washers operate on tap water, the flow rate of which must be within certain limits, set empirically. The container fills and empties repeatedly, by a siphon mechanism.

The pipettes should still be tip upwards to dry; the excess water will
evaporate more easily from fine-bore tips this way. When volumetric pipettes are dry and ready to be replugged, this is a good time to check for any that are broken at the tip or which have a crack or chip at the other end. It is crucial, for reasons of safety, that these are discarded. One of the most common, serious accidents in tissue culture laboratories occurs when a manual pipette hulb is fitted to a volumetric pipette. Using too much force and/or an unsafe method of attaching the bulb, the end of the.pipette may break off and jagged glass cause serious injury to the palm of the hand, with a high probability of severing a tendon. This is especially likely when the glass is already weakened

! wttlh down all work surfaces (if not done daily) for media spills i check under laminar flow cabinet work surface I eltnrrgc water in water baths I lop tr;r humidifier trays in incubators I ltttltirttc condensation from incubator bases I clrcck stocks of routine reagents, plasticware' etc' i ettrgrty aspirator jars as necessary (if not done daily) f lop up liquid nitrogen in cell freezers (or twice weekly)
llt
tttt

y ltl <:heck-list:

by a crack. Tendons can be difficult to repair, leaving the victim with

permanent disubility. All re-usable glunswure will have to be sterilized after washing and drying. 'l'his is one of the topiur covorod in Chapter 2.

I r'k ttttsc and sterilize reverse osmosis unit I r'lrcek water ultrapurification cartridges I t'lrrck whether equipment services/safety tests are due I rlelrost I'reezers (as necessary) I t'slibt'tttc instruments and monitors I rllip down and cleanse/sterilize insides of incubators
2t

Coroline B. Wigley

the laboratory is a murti-user facility, then one or more individuals, perhaps in rotation, should be assigned the task of overseeinjihe runnirrg of the laboratory and ensuring that thJ users abide by an agreeJJoJe otpractice. Examples of good practice include:
flow cabinet after swabbing down the work surface with alcohol and drawing Chloros through the aspirator line o dealing promptly with used glassware etc. to be soaked

If

o clearing away promptly-after finishing sterile work, i.e. vacating raminar

Sterilization
PETER L. ROBERTS

o labelling opened sterile equipment and solutions, for re-use onry by the
same named individual

o keeping good communal records e.g. of shared cell or reagent stocks o checking incubated ceils regularly for early signs of contamination o checking incubators regularly for unused cell cultures and More than in

l,

lntroduction

discarding surplus (or contaminated) stoiks according to approved safety procedures (e.g. after autoclaving, soaking in Chloros, etc.)-

Effectivc sterilization techniques are essential pre-requisites for practising the an ideal environment for the growth l Fle ro-organisms including viruses, and it is only by excluding such agents thft cett* can be successfully cultured and meaningful experimentation can be

lft tlf ccll culture. Cell culture provides

most multi-user raboratories (other than where specific hazards, e.g. radiation, are involved), workers in ihe tissue culture Iaboratory must follow local rules. Failure to do so is perhaps more than usually likely to affect the work of colreagues, as well ur yo.r, own. Shared sterile materials or neglected microbial contamination can iesult in persistent or recurrent probIems for the whole laboratory, which are much easier to pr"u"ntlrrun to cure.

Hfflerl out. In this chapter the various methods for sterilization will

be

Fnglde rcd including some theoretical background but concentrating on practlBel arpccts of their use. For additional background information iee refertFes I -3, Guidance on relevant aspects of good laboratory practice and pod nrtnufacturing practice is given in reference 4 and in Chapter 9.

Acknowledgements
Thanks are due to Peter wigrey for help in preparation of the illustrations and to Andrew Kent for helpful commenti onin" first draft of the manuscript.

1,1 What is sterilization? fre nrethods used for sterilization have a long history. Pasteur first used heat F prcncrve wine (1864), and even before this heat was used in the canning
lnduntry (c. 1700) although with no understanding of the microbial agents lnVolver,l. Other sterilizationtechniques such as filtration were first used in the

lrte lll(ltls.

References
1. 2. J.

Biol. Med., 4, 140.

4. 5. 6.
7.

Bterilization is defined as a process which removes all living things. In pfnelk'c we are really concerned with micro-organisms i.e. protozoa, some filRgl , ltrcteria, mycoplasma, viruses, and unconventional agents [e.g. the Flctltn clusing scrapie and bovine spongiform encephalopathy (BSE)], all of thL.h rrrc invisible to the unaided eye. Although sterility is an absolute term, ifl prncticc methods are used that give ahighprobability that an item is sterile. Elet'llizntion methods vary in their severity and the properties of the item Ferllng to be sterilized also need to be considered. Micro-organisms also vary ln their scnsitivity/suitability for different methods of inactivation or removal.

It. 9,

1,1 'l'he importance of sterility in cell culture

usA,76,
28
574.
1fl3 ot

I0.

Whetr uprplied to a cell culture, sterility means the absence of any contaminatgnnism apart from the cells of interest. Such sterile cultures are essential

htnlE irr rcsearch and diagnostics, and also

for the production of various

Peter L. Roberts

2: Sterilizotion *rElght line. From this the time required to kill any desired number of ffnrrisms can be determined. The efficiency of different sterilization Fgtlrocls can be compared by determining whether large numbers e.g. >1 x l0l, crrn be inactivated or removed and in what time.

biological products. The presence of contamination can have a wide range of deleterious effects. Examples include:

o complete destruction of the cell culture in the worst case o problems and artefacts in research or diagnostic studies o effects on cell-derived products in terms of yield, stability,

etc.

compromised safety of cell-derived products for therapeutic use contamination of other cultures also being handled in the laboratory

t,

Wet heat at up to 100'C

1.3 The use of antibiotics

when a cell culture is free of contaminating micro-organisms, sterility can be maintained by the use of good aseptic technique as outlined in this and other chapters. This represents the first and best line of defence, but the use of antibiotics may also have a role (see Chapters 5 and g).

lQUlpnrcnt and media. Heat can be used in the dry state but is much more lffcctivc in the wet state and is at its most effective at temperatures of 121'C l jt'crrtcr. In this section the use of wet heat at temperatures of up to 100 "C is

flgnl lcpresents one of the most commonly used methods for sterilizing

3[rldcrcd.

t,l

'l'heory

2. Basic principles
Sterilization techniques are designed to kill or remove a wide range of microorganisms including, in descending order of size, the protozoa and fungi, bacteria, mycoplasma, and finally viruses. In all cases ihe agents are composed of single cells, of varying complexity, except for the viruses which are basically composed of only nucleic acid surrounaea uy a protein coat and in some cases a lipid-containing envelope. In addition to these, even simpler agents of disease are known: the so-called'unconventional agents' or prions, which cause slow degenerative encephalopathies in man and u.rimuis e.g. creutzfeldt-Jakob disease, scrapie, and BSE ('mad cow disease'). Howevei, the exact nature of these agents is still controversial; one theoryis that they are infectious proteins (5).

The principal target(s) on which a method of inactivation has its effect varies with the technique and with the micro-organism concerned but may include the nucleic acid, proteins, or membranes. Micro-organisms also vary with regard to their sensitivity to the sterilizing technique 6eing used:

Miny rnicro-organisms, including bacteria and viruses, are relatively easily lfllcllvntod at temperatures of 60-80'C in the wet state. In fact, temperatures ef t.t 60'C are often adequate. However, there are micro-organisms that are FOfc resistant to heat, such as the spore-forming bacteria and the unconventlttttnl ngcnts. In addition, even for any susceptible type of micro-organism, lflUlntrts or variants may exist that are less heat-sensitive. The exact mode of intcllvttion will depend on the micro-organism and conditions involved, but dEltnlur:rtion and coagulation of protein are thought to be important. How!Yl', olhcr targets-such as nucleic acid-may be involved in some situallOnl, Wittt some agents, including the viruses, inactivation kinetics often give fhe to n curve with two (or more) components, e.g. an initial phase of rapid lEtcllvltion followed by a period when the rate of inactivation is lower, This Fltt xuggest that two or more targets are involved in inactivation or that llSt'cgntcs or genetic variation exist within the population. For the highly fEtlrlnrrt bacterial endospores, the target of wet heat is believed to be the
FN A , cc ll membrane, or cell wall precursors and the mechanism of resistance

Itt ltent irrvolves dehydration of the cells. The stability of micro-organisms (lnclrrtling viruses) to heat can be influenced by pH, and by levels of salts,
pt'ttlclrr, irnd sugar. Itt plrrctice the use of wet heat at temperatures of up to 100'C has only a f?w llrrritcd applications in the cell culture laboratory and it is preferable, lhet'c possible, to use other more effective techniques such as autoclaving (FFt'lkrn 4) or dry-heating at high temperature (Section 5) for more effective llct lllzrrtion, Hevcrrrl points should be noted when considering the use of wet heat below

o bacterial endospores are resistant to heat o unconventional agents are resistant to heat, irradiation, and detergent o non-enveloped viruses are resistant to organic solvents and detergents o mycoplnsrnu, viruses, and unconventional agents, are not removed by convonlionnl 0,2 pm storilizing filters

'ltr unclerntanrl lirlly thc cl'ficiency of any sterilization method that depends on lnsetivailon, it ir irnporlunt to know the kinetics of the process for relevant micro-organiamn, pnrlicultrly those that are resistant. A graph of the lognrithm of the numhor ol' surviving organisms against tiile often gives a
28

lllll'(' lirr sterilization:

{B) 'l'lre tlegree of sterilization will depend on temperature (and time). (hl tiporc-ftrrming bacteria and unconventional agents are resistant to all but llrc rnrlst severc conditions, i.e. autoclaving.
2g

Peter L. Roberts (c)

2: Sterilization
?fbtr

It

can be useful for heat-sensitive materials that cannot withstand more

severe heat-treatment.

t.

Standard autoclave cycles

(d) It can be used to inactivate viruses in heat-sensitive products (see below),

Time (min)
lrF
30
15 10 3

Pressure

Survivalb

(barl"
0.7
1.0 1.4

Equivalent timec {min}

60
15

3.2 Pasteurization
Heat-treatment at temperatures of about 60-80'c has long been used in food industry. For example, milk is treated at 63.C for 30 min or at72"C
15 sec. These procedures will kill vegetative micro-organisms and viruses not bacterial spores. Pasteurization at 60"c for 10 h is used in the pharmaceutical industry for certain products derived from human plasma as a met of inactivating any viral contaminants, such as the human immunodeficiency virus and hepatitis B and c, that might be present. Some of these products such as human albumin and transferrin can be used as supplements for cell culture. Before heat-treatment, stabilizers may need to be added to protect the product, but these must not compromise the effectiveness of viraf inactivation. For example, in the case of human albumin, sodium octanoate can be used as a specific stabilizer. Another process that uses wet heat, combined with detergents, is that employed in the washing machines that are used for decontaminating re-usable equipment particularly in hospitals.

t:t

r3

li1
jl har ' los Pa.
lfCr

2.0

in in in in

104 108
1017 1032

4.7 0.8

a relatively heat-resistant bacterial endospore such as BacrTlus stearothermophilus,

ETima required to give endospore survival

of 1 in 108, i.e. time equivalent to

121 "C

for 15 min.

EBgt extreme conditions. Recommended conditions are 132"C for t h in a pfevlty displacement autoclave or 134-138"C for 18 min in a more efficient pFour-load autoclave (2).

tllut the various cycles are not in fact equivalent and that, when survival is mltatcd with the widely used cycle of I21"C for 15 min, other recommended Jlec rrrc either less severe (115"C) or more severe (126" or 134"C). Qneonventional agents are very resistant to heat-treatment and require the
En

3.3 100'C
Boiling or steaming at 100'c for 5-10 min represents a simpre and effecti means of sterilizing equipment and fluids. It is very effective and will ev destroy some bacterial endospores. However, some types of endospore also unconventional agents are resistant to boiling and require more seve conditions for inactivation. It is possible to inactivate endoipores by the u of several 100'C cycles. This method, known as Tyndallization, invol using three cycles of heat-treatment with intervening periods during which spores germinate. This type of procedure can be used in the cell cultu laboratory for decontaminating incubators that have a heat sterilization option

1,3 Sleam
FBf rtelm to be fully effective it must be at its maximum water-holding Epaeity, but with no droplets to wet wrapped items and it must not be lEperltelted i.e. in an over-dry state. This aspect of steam quality can be hgekerl by an autoclave engineer determining the steam-dryness value. The pflenc:c of too much air, i.e. non-condensable gas in the steam supply, will Bllc rrduce the effective steam pressure in the chamber. This too can be Fhcehetl by an autoclave engineer in a non-condensable gas test and may tsBulrc rnodifications to the boiler or pipework for improvement. The presEBee ol'air in the chamber, either from non-condensable gas in the steam SUpply or because of poor air-removal, will lead to a lower temperature in the henrbcr. For instance, if only half the chamber contained steam then, under eEndilirrns of pressure for which a temperature of I2l,"C was expected, only 1lE'(' would actually be reached. If localized pockets of air were to remain, E,g, lrr packaged goods or in infectious laboratory discards, this effect could be Fede r:ven worse. Adequate quantities of steam must be available so that no
RBte llrirn u 10"/' drop in the supply pressure occurs when the machine is operBtlng, ltor pharmaceutical applications, 'clean-steam' is used which involves pfuviding the boiler with purified water of a conductivity of less than 15 pS.

4, Wet heat above 100 "C and autoclaving

4.1 Theory
Boiling water can be a very effective sterilizing agent, but in practice lteverc sterilization methods that are capable of inactivating bacterial endoEpore$ nre requircd for many applications in the laboratory. using water at telnperuture ol'ubout l2loc, i.e. as steam under pressure, is such a method. 'lir proeluee slennr under these conditions special equipment is required i.e. an autoclave. By hnving a knowledge of the inactivation kinetics for reclrtent traeterial errdoepores, suitable autoclave cycles have been deve that give a htgh arnurance of sterility. Some of the standard autoclave tem. perature:time cyelea that nrc commonly used are given in Tahle L It can be
30

!l,B 'l'ypes of autoclave

The type s of autoclave available vary widely in size, complexity, and versatility. Thev rrrnge l'rom small, simple pressure cookers, which generate their own 31

Peter L. Roberts steam, to large machines which are fully automated and may possess multiple cycle options. The nature of the load is important for determining which cycle

2: Stedlization

(ll

is most appropriate.

'Fhey have poor air removal and are therefore not suitable for porous Iontls such as wrapped items or laboratory discards. [b) T't,. use of wet steam and no, or limited, drying capability can lead to wet
Itettts.

4.3.1 Portable bench-top autoclaves i. Upright models

These are essentially pressure cookers as used in the kitchen at home. Steam

is generated in the chamber base, either by an external or internal

heat

source, and an air-steam mixture is discharged from the chamber through hole in the top. A pressure gauge, possibly a temperature gauge, a p control valve, and a safety valve are present. In more sophisticated models there may be an automatic timer, safety features that seal the lid until t

(Cl 'l-trr tcmperature in the load is not monitored. [) Tlr"t" is no cycle record i.e. print-out or trace. (31 'l'he krad only cools slowly after sterilization. [f) Eafety features may be limited or absent. (l) 'l'tre ir use is limited to small items that have not been wrapped and bottled fluirls with loosened caps.

cycle has been completed and the contents have cooled to 80"C, cycle-stage indicators. A guide to the operation of a basic model is given in Protocol L

*t,l
ilhlle

Protocol 1. Operation of a basic upright bench-top autoclave

1. Consult manufacturer's instructions before using instrument. The following is for outline guidance only. 2. Remove lid and fill to required depth with water. 3. Place the items to be sterilized into the machine on shelves or in a
basket.

100 litres) floor-standing which are either top or front loading, these are often essentially l$lttglavcs, hflet vcrsions of the bench-top models already described, although more lFhlrlicltcd types exist. The next major advance in design involves the use Ff iR external steam supply. In these, air is displaced from the bottom of the henfhcr by the less dense steam entering at the top. These machines also Stllerr rr iacket for water or steam, or spray-cooling systems in some bottledF$ldr nulrrclaves, and thus cool the chamber contents more rapidly. The use f the .inckct, combined with a simple vacuum system, can be used to aid in ihc tttvi"g of the load. This type oi autoclave is:

Gravity displacement arrtoclaves tlrcre exists a range of larger capacity (e.g.

4. Replace lid and seal; heat water to boiling. 5. Allow steam to escape for several minutes before closing the pressure
valve (if appropriate).

i I I

Urefrrl

lirr items that are easily exposed to steam, e.g. glass or plasticware

6. Start timer when the required pressure and temperature are reached. 7. At cycle end, turn off heat and allow pressure and temperature to fall. 8. Open the steam valves and allow to cool to 80 "C and atmospheric
pressure before removing the contents.

rultsble ltrr bottled fluids llttl rruilirble for porous loads and wrapped items because the air removal ffielllrrl is not adequate

ftt,S Multicycle porous-load

autoclaves

ii.

Horizontol models

These are essentially similar to the upright models described above but h more features such as cycles at both lzl"C and 134"C, a drying option heat (but not vacuum), safety features such as a temperature lock, indicators, and automatic timers. However, steam is generated in an ident

way to thc upright models.

iii.

Usee

cnd linritotions

Somc of the poinln thnl nccd to be considered when using these small types uutoclnvo nro: 32

?gftttts krird autoclaves were developed because of the problem of air reHttval lront wrapped items and other loads; these use a vacuum which, when Hferl lrr prrlses combined with steam, efficiently removes air from the load and Ehanther'. lt thus helps the penetration of steam into the load. The use of a lr,rrrriir rrlso aids in the final drying stage. Such machines tend to be large, illllt a rrrinimum overall size of about 4 m3 and a chamber of at least 0.5 m3, 6fltl expr:lrsive. Some machines used in hospitals or the pharmaceutical ind1tlty ,,r'e far larger and materials can be directly wheeled in, or loaded via detllt,trtetl trolley systems. In the laboratory a multicycle machine is commonly Uterl, willr ir number of different cycles (Frgure 1) available for use depending Ell llrc rrrrlrrrc of the load; this increases versatility in a multi-user situation. FHlntttt.lcrs such as temperature and time can be pre-programmed into the Fulrirrt: rrnd are often not user-adjustable. In addition, a cycle may be YHllrrtrle irr which any comhination of these parameters can be set by the user

lts

Peter L. Roberts

2: Stefilization
Drying

Air-removal Sterilizing

A+
o

{bl Dil'l'erent cycles are available allowing flexibility of use. te) 1'he chamber has a large capacity e.g. )0.5 m3, and will accommodate ,!1. 40 x 500 ml bottles, several large 5 or L0 litre vessels, or bags of
dlncards.

Sterilized items are free of moisture due to the presence of a vacuum drying stage at the end of the cycle.
Eeenuse of the wide range of options and cycles available, this type of Eptgcluvc is usually made to order. Thus the user must carefully consider the

Sppllcntions and loads for which the machine will be used, *th a sterilization engineer and the manufacturer.

in consultation

l,*
C

Preparation of the load and operation of the machine

+ o

futat\il

bgstc outline procedure for preparing material to be autoclaved is given in 2.

Ffotocol
D + o

2.

Preparation of the autoclave load

A, Equipment

t,

t, l,

Wrap items in aluminium foil or place in sterilization bags using autoclave tape. Do not seal completely. Loosen the caps of bottles and other containers, or wrap up the open
tops.

Figure 1. Typical cycles found on a representative multicycle porous-load

Pressures within the machine, whether positive (+) or negative

(-).

are indicated. Cycle

which involves both negative and positive pulsing, is used for fabrics, assembled units, and discard loads from which it is particularly difficult to remove air. Cycle B uses single negative pulse followed by positive pulsing and is used for laboratory pl and glassware. Cycle C, which uses a series of negative pulses, can be used for d loads. However, in validation studies, cycle A has proved more effective for discard loads. Cycle D is a bottled-fluid cycle which uses air displacement to remove

l, t, l.

$ome items, e.g. pipettes or micropipette tips, may conveniently be placed in containers such as tins or beakers which are then covered with aluminium foil or a sterilization bag. Looseh the tops of tins. Micropipette tips are best sterilized in the boxed racks supplied by the
manufacturer. The necks of autoclave bags containing mixed discard loads must be lrlosened to ensure good steam penetration. Add a piece of autoclave-sensitive tape to the items in the load.

immediately before use. Pressure and temperature during the run are cated by dials or digital meters. In addition a printer and/or chart-recorder fitted to allow the temperature throughout the cycle to be recorded. A rccordcr is morc easily checked by the user. On more recent models an a detcctor is usuully ['itted to confirm that all the air has been extracted from t chantber. Some points to consider with a porous-load autoclave are:

H, Fluids

L For a fluid cycle, bottles may be either sealed or left unsealed. L e heck caps for their ability to withstand autoclaving. With some types I,
of cap the rubber liner may be sucked into the bottle. Add autoclave-sensitive tape to the items in the load.
use, the sterilization of typical loads should be validated as ibed in Section 4.5, The heat-scnsitivity of the items should be considered 35

(a) Thero
anel

in goocl rtteflm penetrution of porous loads, allowing wrapped

[lrlirlc routine
cr,r

ditficult itemn to be sterilized

34

Peter L. Roberts styrene cannot. Fluids such as water, salt sorutions, and buffers can also autoclaved. The heat-sensitivity of more complex formulations containi components such as vitamins, growth factors, antibiotics, or heat-sensi amino acids, e.g. glutamine, should be considered or tested. cell culr

2: Sterilization

and tested before use. Many plastics can withstand 12roc, but

I, t.
10,

On completion the cycle-end indicator will come on.

Check the chart paper or print-out for a satisfactory run.

Open the door and remove items using heat-resistant gloves to prevent injury. Check autoclave tape or other in-load indicators. Label each item to indicate sterility if autoclave tape has not been applied to every item.

lr, tt.

Complete log book record.

l,i
use on every item allows a clear distinction between items that have sterilized and those that have not. Such indicators are essential with the basic types of autoclave considered earlier which do not record details of cycle' The simplest system is the use of autoclave-sensitive tape, but ar clave tubes [Browne's tubes (solmedia)] or indicator strips can be used. So types of autoclave bag possess integral indicators. only an approxim indication that sterilizing conditions liave been met is given by zuch indica tors. Nevertheless, they do provide a measure of assurance. Ii there is an, problem with the wetness of porous loads at the end of the cycle, conside tion should be given to modifying the length of the dry-cooling stage or sim to repeating this stage at the end of the run. proiocolj o-uttiries trow autoclave should be operated.

Autoclave testing

Fgr nrr autoclave to perform satisfactorily and to ensure the sterility of the lgltl, it is necessary for tests to be carried out by the user and by the lfrlllzution engineer, and for records to be kept. This is also essential in fdet to meet good laboratory practice and/or good manufacturing practice $ldelincs (see Chapter 9). Such testing goes hand in hand with a regular pflverrtive maintenance programme. An indication of the sort of tests that ihrrulrl lrc performed (6) is given in Protocol 4.

llotocol 4. User tests and records for the autoclave

l, l, f,

Hocord details

of all runs, including pre-use warm-up and leak-rate

tEcts, in a log book with details of date, load, cycle, result (pass or fail),

oprator, and other relevant information.

Protocol 3. operation of a murti-cycre porous-load autocrave 1. Follow the manufacturer's instructions and standard operating procedures.

Carry out leak-rate test, preceded by a warm-up run, at least weekly.

At the end of every run, check the chart and/or digital recorder for

2. carry out routine or weekry

filter.

tasks such as cleaning the door-sear, removing debris/broken grass from the chamber, and creaning the drain
(see

ratisfactory cycle. Relevant parameters include temperature, time, and lrrrlsing profiles.

3. Perform warm-up, leak-rate, or performance tests as necessary

Protocol
4l-.

At regular intervals, e.g. daily or weekly, carry out a Bowie-Dick test laae Protocol 5l or similar test to assess steam penetration into a ctandardized difficult load (7).
Arrange for a sterilization engineer to carry out servicing at regular Irrtervals, e.g. every 3-6 months, and keep records of test results and work performed. llofore the machine is first used and at regular intervals thereafter, e.g. tl nronthly, arrange for the sterilization of standard/typical loads to be valldated by thermocouple studies carried out by sterilization/quality
eeBUranco personnel.

4. Load the chamber but do not overfill. 5. Fill in dotails of run in the log book (see protocol 4il. 6. Place the wander-probe into the load. ln the case of a bottled-fluids
7. Scloot approprlatc cycle and start the run.

cyclo, place tho probo into an identical control bottle with liquid which cen bc dltcardod later.

97

Peter L. Roberts

2: Sterilization

Protocol 5. Bowie-Dick test for porous_load autoclaves

towels is both rapid and even (7). lt is performed after an initial warm_up
run.

!, Dry heat
Hfntlng in the dry state can also be used for sterilization; however, because thh lt not as effective as wet heat, higher temperatures and longer times must bi uncU. Under these conditions the inactivation of micro-organisms is pflmarity due to oxidation.

This test is designed to confirm that steam penetration into a test pack of

1. Fold cotton huckaback towels (g50 x 500 mm) in half three times.
These towels should be clean, dry, and not over_compressed. Open them and allow to dry between tests.
Make a test-pack from the forded towers to give a height of about 270 mm (25-36 towers). rf more than 36 towers are required to reach this

2. 3.

height, dispose of the old towels and use new ones. no.

12221 in the centre of the pack. The tape must not be more than 1 year past the manufacturer's date stamp. speciar ready-prepared test

Place a sheet of A4 paper with a cross made of autoclave tape (e.g. 3M

Aftho(l lbr destroying micro-organisms. For instance, during routine aseptic ffhniquc, the openings of glass bottles and other containers may be briefly tfrfmett' i.e. placed in the heat of a Bunsen burner (although this may not be fiffmnrcnded when working in a laminar flow cabinet-see Chapter 4). This Ei lle carried out even with plastic flasks and tubes if care is taken to keep

Incineration * ccll culture laboratory there are a few examples of the use of this ln
tt

trl

sheets can be used.

4'

Tape the pack together without compressing it further and/or prace it in a wire basket to prevent it failing over. place it carefuily in the chamber on the longitudinal centre line and c. 100 mm above the chamber floor, and carry out a standard porous_load cycle.

S! ex;t,rrur" time to a minimum. In the microbiology laboratory a metal loop b ute.l l'or transferring cultures and this must be completely sterilized beHicn usos. This is carried out by heating to red-heat in a Bunsen flame. Tltel itcms and scissors, used for example during the preparation of primary tsll eultures, may simply be re-sterilized between manipulations by being
Sppe,t irr cthanol which is then ignited and allowed to burn off.
lneltrcration is often used for the safe disposal of laboratory waste. However, *hlle this is an effective method for use with microbiologically contaminated

at steririzing temperature was within the specified limits for the cycle used. 6. check the autocrave tape. rf the machine has been operating correctry, the colour change arong the tape wiil be even. rf an area wiih a righter colour is found, this indicates a cord-spot where poor air-removar has occurred' An autocrave engineer must be contacted and the machine must not be used until the fault has been rectified.

5. At the end of the run, check that the hord-time

Ftterlnl, it is recommended that such material should be autoclaved or Hfted hy chemical methods in the worker's own department prior to incinerItlCn, 't'hus any problems due to transportation, inefficient incinerators, or
inndeepr,,t" direct supervision are avoided. Incineration may be an acceptable

During validation of standard/typicar roads, the parameters that should tested, and the limits expected, include:

(a) porous-load thermocouple

tests

ffpo*rt route for routine cell culture materials but, if contamination is known f tscltcctcd e.g. in accidentally contaminated cultures, or cell lines transFfmetl with viruses, the above two-step procedure should be carried out. For EF lncllr(.rator to operate efficiently it should reach a temperature of 350"C Bfld ptclcrably have after-burners fitted to incinerate any unburnt material lhel rrrr,y bc found in the smoke.

!'fl !kll-air ovens

ii'
i.
other relevant loads should be tested including worst-case/difficu loads and items e.g. discards, filters, and wrappJd items.
in the dry state provides a commonly used alternative to the for heat-sterilization. Some points to consider when using this HIE of Fethorl itrclude the following: {al lt ix sirnpler and cheaper than using a porous-load autoclave but longer Irrrrllncnt times and higher temperatures are required e.g. > t h at 170'C (erelutling load warm-up). lb) lt"rrrs do not get wet during the process. [el !t in nrutincly used with glassware and metal objects.
Hentirrp. itcms
nlc:irm

(h) liquid loud rcsrs

shoukl ilgrce lo within t.C,

'l'h-crmocouplc tests on bottled fluids should show ll5_117"c for a ll.5nc'cycle, ur rzr-rz4'c for a'r}'roc cycre. Ail probes in the road
30

3g

r
Peter L. Roberts

2: Sterilizution
e

(d) some plasticware can be treated but requires a lower temperature I?O"C for > 18 h.

plasticware.

A hot-air oven comprises an insulated box with electrical heaters. Heat distributed by simple convection or more effectively by a fan. For mo critical applications, e-g. in the pharmaceutical industry, u Rtt", may be to the air vent to prevent any possible recontaminaiion of the load occurring during cooling. The required temperature and time is set and t actual temperature monitored by an inbuilt digitat or dial thermometer ar or a thermometer placed through a port into the chamber. Some timers ar only triggered when the required temperature is reached in the chamber. T monitor the temperature continuously a chart-recorder may be included. F large volume sterilization on an industrial scale, tunnel-sterilizers are used which items are continuously fed by conveyor through a hot-air tunnel. various temperature-time combinations can be r^"d fo, sterilization. most commonly used conditions, and those generally recognized by regulatory authorities that control medical products, are-160 "cio r 2h, i7o for t h, and 180"c for 0.5 h. These excludelhe time required for the oven a load to warm up to the required temperature. other conditions can be ur where items cannot withstand these high temperatures e.g. 150.c for 2.5 140'c for 3 h, or r20"c for trg h. The latter conditi6ns are useful f sterilizing certain types.of plastic, e.g. poryallomer and polycarbonate ( used in centrifuge tubes) or polypropylene (as used in micropipette tips boxes). However, they are not suitibie for porystyrene. the'supplier,s r logue should be consulted for further detaili on ttt" heat-sensitivity of

f,

l6fety device which prevents the door being opened until the temperature has dropped to a preset level. Check chart and in.load sterilization indicators for a satisfactory cycle !nd then remove the load. Mark all items as sterile.

Routine monitoring and testing

indicators can be used as simple checks to confirm that the load has heut-treated. Because most of these are just indicators of the temperature nnd not the time, they do not indicate sterility. However, they do ru useful indication as to which items have been in the oven when all items ilied with heat-sensitive tape. The most commonly used and cheapest is the use of heat-sensitive tape on which dark stripes appear when
oC are reached. Alternatively, indicatorstripswhich change falures of 160

ovcr the range of e.g. 116-154'C or 160-199oC, are available. rure less commonly subjected to extensive routine validation comwllh autoclaves, partly because they are much simpler to operate and llttle or no routine servicing. However, official standards do exist and it blc to test the performance and instrumentation of an oven after purchase and on a regular basis, particularly where an oven is for turc in the hospital or pharmaceutical area. Checks should include:

I 'I

Ullng t hermocouples placed in contact with the oven temperature indicator

5.3 Load preparation and oven use

The basic steps involved in preparing items for oven sterilization and carrying out the process are given in protocol 6.
Protocol 6. Sterilization using a hot-air oven
have their opening wrapped or capped (check heat-sensitivity of the cap). 2. Add heat-sensitive tape, or other temperature indicator, preferabry to every item in the load.

I rlpplc <1'C I drlll <2"C

It

warnr-up <135 min ovcrshoot <2oC

Eeattlngs of these terms are illustrated in Figure 2.

lgre nhould be within

Jrrtlgcd by thermocouples placed around the load, the load tempera-

oC

of the oven temperature indicator

1' wrap items in aluminium foir. containers need onry

Irrrrrliation

,*-1 Ultroviolet light

.X$tfntl,tto, (tJV) light at a wavelength of about 260 nm will inactivate micro'Ei$nlurn ulrcl viruses. It acts on nucleic acids, leading to strand breakage, Fnd elorr-linkage, and the formation of pyrimidine dimers and other prodl$l, H,rwevcr, because of its poor penetrating power, its usefulness is very Fftlteet ll cnrr be used to sterilize air in cell culture cabinets and rooms, and rl'ler they have been cleaned. As the power of UV lamps decreases
Ute, I lrey should be checked on
a

3.

Fill the oven by placing items on shelves. To aid circulation of air and to promote warming up, do not over-fill.

4. sat tlmer, temperature,

tlm to allow for oven and/or load warm_up as necessary. 5. On complrtlon, allow ovsn to cool. Some ovens have an automatic
40

and chart-recorder controls. Add on additional

regular basis. As found with heat, bacterial

glttl unconventional agents are highly resistant. Also many organisms l)NA repuir mechanisms that can overcome limited damage.

4l

Peter L. Roberts

2: Sterilizqtion
p*ldc. 'l'heir activity is greatest at higher temperatures and humidity levels of 75::1tttl'U,. However, conditions that reduce the accessibility of the micro,tgenisrns to the gas, e.g. being dried in organic or inorganic material, will defeu*" the effectiveness of fumigation. These fumigants act as alkylating EEgnts and act on both the nucleic acid and protein components of the microfganisrns.

W
o E o
CL

-Ripple

Overshoot_

1t1,1 Iithylene oxide thyl.t," oxide is commonly used for sterilizing items of clean equipment in lry=terrrperature autoclaves, or in combination with steam, particularly in hglpltutr. In the laboratory some items of plasticware may be purchased preIttillzect with ethylene oxide e.g. syringes and filters. However, this method
Warm-up Sterilization Time

Cool{own

9f eterilization can leave behind toxic residues, and other methods,

e.g.

lEmrna-irradiation, may be preferable (see Section 8.1.1). Some additional plntt to consider include the following: @] Ttre gas is toxic and operator exposure needs to be controlled. [-b] Speciat equipment is needed. Gl tt i* not suitable for decontaminating bulk discard loads.

Figure 2. Typical cycle for a hot-air oven. points in the cycle (see section 5.4) which important in confirming the effective operation and validation of an oven are indica The ripple effect has been magnified for illustrative purposes.

6.2 Gamma rays

Gamma-irradiation, although not actually performed in the laboratory, is very important method of sterilization. It mainly acts directly on nucle acids, but indirect effects by free radicals and hydrogen peroxideiormed water may also be important. Unlike UV light, it has very good penetrati properties. It is commonly used with items that cannot be heat-steiilized, a has the advantage that, because of its good penetrating power, items can completely sealed and packaged. various items of disposable plasticware bought pre-sterilized in this way e.g. tissue culture flisks and dishes, filte plastic pipettes, syringes, pipette tips, and some chemicals such as antibioti such items will have been expose d to 2.5 Mrad, the accepted standard r in a cobalt-60 plant. This exposure level has been in use by industry for a period and has been found to be acceptable as long as only low level llmination is present which is not unusually resistant. Bactlrial spores viruscs lend to be somewhat more resistant to irradiation and unconventi lrgcnt$ nrc cxtrcmely resistant,

ir1,l
ffilf

lrrlrmaldehyde gas

gu* is commonly used to decontaminate or sterilize larninar flow cabinets

13 ro,rrns used for handling cell cultures, and also for small items of equip-

EEat, A'r outline of the procedure used when treating roorhs is given in fuAatl7. It should be noted that formaldehyde gas is toxic and thus any Eerrary precautions, including the use of breathing apparatus, should be *Ftn *lr.rr" necessary. Frotocol 7. Fumigation of rooms using formaldehyde.
llernove all unnecessary items from the room, including those that nrust not be exposed to the gas e.g. cell cultures, sensitive electrical er;tripment, etc, (ilcan the room to minimize the level of microbial contamination and lo allow good gas penetration.
as lor as possible.
itiace a hot-plate with large saucepan, connected to a timer, in the room (tlrotime requiredto boiloffthe liquid should bedetermined in advance). Altornativoly use a dedicated formaldehyde-generating kettle."
E.

t,

t, Irrrn off air-handling system and seal up the room with masking tape

7, []homical sterilization
7.1 Fumlgation
Formaldehyele gan trr er.hylcnc oxide can be used for fumigation. Both el'l'ective againrt ull types ol' micro-organisms including u'irur"*, alt bncterial endoeporen ule s'mewhat more resistant in the case of ethy
a

I iii tlrc container with formalin solution (20 ml per m3 of room volume),

!, Irrrn on ths hot-plate or kettle and loavo tho room.

4S

Peter L. Roberts Pro_tocol

2: Stefilizstion
attach
Trblo 2. Effect of liquid disinfectants dn micro-organisms Effectiveness"

7. Continued

7. Lock the door, tape it up fully (including the keyhole), and

safety warning notice. 8. Leave until the next day. 9. lf a total exhaust cabinet or air extract is present, turn this on by a remote switch. lf this is not possible enter wearing breathing apparatus and turn on cabinet and/or air-handling system. 10. Leave until the level of formaldehyde reaches an acceptable level. Testing equipment should bd used. 11., The room may require cleaning to rerno-ve residues of paraformaldehyde and it may take several days to remove the gas completely,
" Other methods for generating formaldehyde gas exist e.g- heating paraformaldehyde (10 g/m3) ' and, for those situations where fumigation is carried out regularly such as in pha(rnaceqtical manufacturing areas, equipmenl that generates a.mist of formaldehyde (e.9. Phagojet, Labor-, atoires Phagogene).

Typc Aldehydes Hypochlorites Fhonolics

Aloohol

Fungi

Bacteria Endospores
+ + + +

Viruses

+
+ +

++ ++ -

+lv +/v

tEflective (+) and non-effective (-). ln the case of non-enveloped viruses, the 3lleot may be variable or partial (vl depending on the particular virus. Examples of tht dlfferent types of disinfectant are given in Section 7.2'

Table 2. Disinfectants can either be prepared directly using ry reagents or be purchased in the form of proprietary formulatidns.

in

6f the factors to be considered when using or selecting a suitable

Microbiological safety cabinets (only models that can be sealed and to the outside) that are used for handling cell cultures may also be taminated with formaldehyde (Protocol S) on a regular basis, e.g. once week or once a month, or after it has been,found that contaminated cultu
have been handled. ant are listed below.

ffhe rnnge and level of antimicrobial activity is less than with other Egthods of sterilization, e.g. heat. Bpre-forming bacteria and some types of viruses may be resistant.
SEme disinfectants are neutralized by organic matter.

Protocol 8. Fumigation of laminar flow cabinets 1. Clean the cabinet. 2. Fill the integral formaldehyde generator attached to the exterior, or a
small portable generator placed inside, with c. 25 ml of formalin solution.

atnbility of 'working' dilutions varies with the type of disinfectant. ctposure time required depends on the type of disinfectant and on

it is used. ty to the user should be considered. disinfectants can be used in the cell culture laboratory for:
Utlne hygiene and disinfection of items of equipment and surfaces in

3. Also place any items of equipment

cabinet e.g. pipette-aids etc.

used in cell culture procedures in the

4. Replace the door and seal with masking tape. 5. Switch on formaldehyde generator and place a warning notice on the
cabinet.

:fi llii:T:?:*

6.

Leave overnight before turning on air-exhaust while opening cabinet

curture items, e.g. grassware, arter use and prior to und re-sterilization *e:hing tfCatrnent of used or contaminated cell culture media before disposal

door.

7. Leave to vent before cleaning cabinet to remove residues of

paraformaldehyde,

7.2 Liquid disinfectants

Many types of liquld disinfcctant exist and they have a useful role in the culturc laboratory, The mnin properties of some of the principal types
44

most suitable type of disinfectant for a particular application varies below;, 'l'here are a number of methods available for applying the disEat, Thcse include using a cloth or hand-held sprayer, or the use of towels applied to liquid spills and then soaked in disinfectants. For a !tet' dirinf'ection of a room a 'knapsack' sprayer, or large volume garden willr n long lance should be used. In this case the necessary safety prerhould be taken to avoid exposure of the face, eyes, and lungs. Items re or plasticware are best treated by being fully immersed in the
11t
,

4l

Peter L. Roberts

2: Sterilization
rposure time is sufficient. The ethanol/water mixture is then left to naturally. Although a convenient and easy disinfectant to use, and toxicity, it is not very effective against fungi, bacterial endospores, or loped viruses. It is best used as a cleaning agent or disinfectant for fltical applications e.g. for treating microbiological safety cabinets beEnd after routine cell culture work (see Protocol 10).

7.2.1 Aldehydes
Formaldehyde and glutaraldehyde can both be used as liquid disinfectan
although glutaraldehyde is more commonly used and is probably more ive. They both have the advantage that they are not influenced by t presence of high levels of organic matter and they inactivate all types micro-organisms including bacterial endospores. Formaldehyde is used about 4o/o, conveniently prepared by diluting 40"/". formaldehyde soluti (formalin). Glutaraldehyde is used at2/". Examples of commercial tions include Cidex (Genesis Service Ltd) and Gigasept (Sterling Medicare Treatment times are typically of the order of at least 30 min, with longer ti required to obtain full sporicidal activity. Proprietary glutaraldehydedisinfectants need the addition of an activator before use and then have recommended life-time of about L week. Aldehyde vapours are conside relatively toxic due to their ability to sensitize, and they also have muta and carcinogenic properties. Steps to limit exposure should therefore taken.

Others
other disinfectants that can be prepared in the laboratory or purchased EBftmcrcial formulations may have a place in the cell culture laboratory. include hydrogen peroxide at 5-10"h, acids, alkalis, ethanol (70'/.)

with 4"/" formaldehyde or 2000 p.p.m 'hypochlorite, or Mikrozid nol/aldehyde mixture, Sterling Medicare). One proprietary nt (Virkon, Antec International), that has been shown to inactivate

ng agent (peroxide), acid

rnnge of viruses and other micro-organisms, has three components: an (pH 2.6), and detergent. With this agent the conditions are a1.."/" solution and an exposure time of at least

7.2.2 Hypochlorite
One of the main advantages of using hypochlorite is its relatively low cost a ready availability e.g. in the form of household bleach or Chloros Chemical Distribution Ltd). However, it has the disadvantage that it is not effective in the presence of high levels of organic matter. It also metals and thus must not be used with centrifuge rotors or cell cul cabinets. It is not very stable after dilution. The concentration recomme for use when high levels of organic matter are present is 10000 p.p.m. available chlorine, although lower concentrations have been recomme for routine hygiene e.g. 2500 p.p.m. The concentrate is stable but di disinfectant should be changed after 24 h. An exposure time of at I 30 min, and preferably overnight, is recommended. For treating sodium dichloroisocyanurate powder e.g. Haz-Tab granules (Guest Ltd) can be used to release high levels of chlorine rapidly.

Flltration
Fllters for bacteria and fungi
rlicst examples of this type of filter were first developed in the late end used such materials as unglazed porcelain, diatomaceous earth, or sintered glass. These materials acted as depth filters in which were trapped within the filter. For sterile filtration, such filters have becn replaced by 0.2 pm membrane filters, first developed in the , 'l'hey act more like sieves which trap the bacteria on their surface, gh the distinction between the two types of filter is not absolute. rne filtration is the method commonly used for solutions that cannot lterlllzett by methods such as autoclaving, because they contain heat-labile ts. It suffers, however, from the disadvantage that, because steril. rkrcs not take place in the final sealed container as is the case for FBmpte with autoclaving, steps must be taken to prevent recontamination EEFfee,t liltration and bottling. Uses in the cell culture laboratory may include rrg culture media, sera and other cell culture supplements, and any 'rtl products made by the cells.

7.2.3 Phenolics
Clear phenolic disinfectants, e.g. Hycolin (William Pearson I td), al not inactivated by organic matter, have little activity against bacterial e rtporcs. 'Ihey are commonly used at a concentration of 2-5"/o, or as mcnded by the manufacturer. They also may leave sticky residues when

for cleaning surf$ccs,

7.2.{ Alsohol
Ethenol ls eommonly uned tbr disinfecting surfaces and hands (pre gloved, but aomo peoplc treut their hands directly). Its effect is optimal concentratlon of 7(l:8(lol' flnd surfaces must be fully saturated to ensure t
40

tr1i1 'l'ypee of filter $ rangr ol' filter materials is available which can be used for

removing

Egterln rrrrd fungi (8). For the filtration of liquids, these include materials pp6e trv casting processes such as cellulose acetate, cellulose nitrate, or a

Fi*tute ol'both, or nylon or polysulphono, Other materials with low proteinSadtng pluperties, e,g. polyvinylidene difluoridc, are also available although
47

Peter L. Roberts

2: Sterilization
Egille (i.c. bottle-top filters), some with integral receiving vessel, are also
Blallnlrlc.

for most applications in the cell culture laboratory (such as filtering culture media) this property is probably not really necessary. In additi filters can be made from polycarbonate by the irradiation-etch me These have discrete pores with a much narrower range of sizes and act exclusively by retaining micro-organisms on their surface. Full det on the various types of filters that are available, together with appli
and selection guides and details on chemical compatibility and other

Flltcr units comprising the filter housing and the filter itself can be pur*aletl cither as pre-assembled, sterilized, and disposable units or as a rerequired filter(s), can be sterilized by l*Rble lu,using which, after fitting the systems are il$teeluuittg. Some points to consider with regard to the different Etetl bcl.rw.

can be found in the extensive literature produced by the various m facturers such as Gelman Sciences, Nucleopore (Costar), Sartorius,

[El dlnlt,tsaule units:

Millipore, etc. Membrane filters come in a range of pore sizes, but 0.2 pm is consi the standard for removing bacteria and fungi. Larger pore sizes, and pre-filters e.g. made of glass-fibre, are useful in serial filtration systems increase the filtration capacity where significant levels of insoluble may exist, e.g. in serum, complete medium containing serum, or cell cul supernatant harvests. For effectively removing mycoplasmas, filters with pore size of 0.1 pm are required. Filters with pores of this size are now by most commercial processors of serum products. In addition, filtration this level may also be useful for removing some of the larger viruses that be present in such products, e.g. infectious bovine rhinotracheitis virus
parainfluenza-3.

ltrtr convenient lrnvc

O ltttvc high

unit cost quitity control of filter batches, including integrity, carried out by

lltc tnanufacturer

lbl t='

re"rrsrrbte systems:

t I t

t,ctpire assembly and sterilization before use, followed by disassembly rntl cleaning after use utny a greater risk of failure due to improper assembly nl'lcr initial purchase of the housing, have low unit costs

&19 l'rnr:tical asPects

Low levels of various extractable materials may be found in me filters, e.g. surfactant used during manufacture or residues of the e oxide gas that is used by some manufacturers as a sterilizing agent (9,10). critical cell culture applications, where the complete absence of such is necessary, filters with very low levels of extractable material which been sterilized with gamma-radiation, can be used. Alternatively, practical, simply discard the first sample of filtrate. Apart from filteri liquid, membrane filters are also used. for filtering gases e.g. CO2 or supplied to cell cultures growing in mass culture systems or via a filter inse in the cap of cell culture flasks (e.g. Bibby Sterilin, Life Technologies Ltd, Costar). Various types of filter, ranging from membrane filters to si cotton wool plugs, are also used to protect liquid handling equipment such pipettes, automatic pipette handling devices, and micropipette tips, from or liquid-borne contamination.

e.g. using a l{emhr,,,," filtration can be carried out under positive pressure to c. 50 ml, air pressure from a pump or a SFngn lirf small volumes up fulitrc-tirr", or a peristaltic pump. Negative pressure can also be used with Ftb=t,rl liltcrs and filter units with integral receiving vessels. However, this *-thrxt irls the disadvantage that, with tissue culture medium containing that caused by the Ffh'rrrrrc,, the pH rise caused by filtration is greater than gi*ttlVn lrrcssure methods. In addition, when protein is present, frothing and *leltt ilcrrilturation can occur. After filtration, the liquid can be collected ff*et lrrr.t ir single final sterile container or, where smaller aliquots are *gttn,l, tlircctly into vessels of a convenient size e.g. 500 ml bottles for basal plasticware manufacturers EEll utrlt,,'.' tncdium. The filter, cell culture, and filtration systems designed for use with particuvarious complete

$n ;truvittc

let lirprt,t volumes. llt'l\tiltxtil 9 the typical steps involved in liquid filtration are given.

tt.1.2 "l'ypes of filtration unit Irr ucklilion to thc wide range of filter types that exist, filter housings crtttte itt ll lnng,c ol' shapes, sizes, and materials. The filter manufactu nlruultl hc colrsrrllcrl lilr dctails of the physical and chemical properties
lroth lillclr nnrl lrorrsings. Sizes and formats range from the simple syri lllteL ol' ll ot 25 rtrrrr tlirnrctcr which will filter small volumes, to ca ol'vnrioulr xizes lrrtrl rltrpes, ol'lcn pleated to maximize filter area, for irrg lurga volunror, l)iupottblc units designed for filtration clirect into

Peter L. Roberts

2: Sterilization
I

Protocol 9. Continued 3. Assemble the unit taking care to install all supports and o-rings in their
correct positions.

4. Attach tubing, if necessary, to inlet/outlet ports. 5. wrap the items and sterilize by autoclaving at 12r.c for r5 min using
not be used unless recommended by the filter manufacturer.
B. Filtration

i
I
a

'l'hc pressure-hold test, which is based on measuring pressure decay after pressurizing the upstream of the filter. The test measures not only the lntcgrity of the filter but also that of the filter housing. It is usually carried out using test kits provided by the manufacturers, which can often also porform the diffusion air flow/forward flow method.

porous-load cycre. Higher temperatures, or dry-heat sterilization, must

In atldition to these physical tests, sterility testing of the final product, by

lating a range of microbiological growth media, can be carried out. Both y testing and sterility testing are routinely carried out in the pharmaindustry.

niques, and assemble the complete filtration system. 3. Tighten all connections and attach to pressure system.

2. Remove the sterirized firter from its packaging using aseptic

1. Assemble any accessories that are required in a raminar frow cabinet.

tech-

Fllters for viruses

0.1 pm membrane filters can be useful for removing large viruses,

rly when several are used in series, membrane filters of a much

pore size are required for the removal of most other viruses. Recently

system via the air vent (if present). 5. Test the integrity of the filter unit and/or carry out sterility tests on the

4. Start filtering into sterile container(s), after having bled off any air in the
filtered liquid as necessary.

llltcr manufacturers have introduced filters which fulfil this requireI Millipore (L7,12\, Asahi Chemicals (13), and Pall Corporation (14).
now filters are available in a range of formats, e.g. tangential flow units

8.1.4 Testing of filters

To ensure the correct functioning of the filtration system, the assembled r i.e. filter and. housing, needs to be tested for intigrity. The manufactu :ayl- out various physical tests that are related to ttre actual pore size bubble-point or air-flow/diffusion tests. These are then correlated with actual performance of the filter-or filter unit in tests designed to challenge filters with high levers of a smail test bacteriurl, e.g. psiudomoni dimiru, for a 0.2 pm filter. A number of quantitatiue pro"edures can be carried out the user to confirm the integrity of the filter unit. The tests shouid be ca out after and, where considered necessary, before using the filter. ihe sim method is to confirm that significant resistance (i.e. b-ubble-poi"rpr".r" see below) is felt when trying to force air forwards or backwards throu wet filter- If air passes through unhindered then the filter is damaged. quantitative methods are:

70k or 180k, Millipore), hollow-fibre cartridges for tangential or filtration (Planoray 15, 35, 40, or 75 nm, Asahi Chemicals) or deadflhcr units (0.04 pm Nylon 66, Pall Corporation). All thcse systems use membrane filter technology except for the hollowiy$tcm which acts as a high-performance depth filter. The Viresolve l'rom Millipore and the Planoray system from Asahi have been the fully validated of this new technology. The Viresolve system has been to remove viruses exclusively in a size-dependent fashion. Some filter eolnc in a range of pore sizes and the most appropriate type that will ffilze virus removal without removing significant levels of any important nrnrponent or product should be selected. For instance the pore size of lerpcst Millipore Viresolve membrane has a nominal molecular weight lrrting of L80,000 Da which is close to the molecular weight of imnughrbulin G (IgG). It is thus essential to test the suitability of such filters
nny spccific application. Fttme rrspccts to be considered

with regard to virus-removing filters are

lbt:tl

hckrw.

(a) 'l'he (b)

.bubble'point method, which is based on determining the pre requirccl to force air through the wet filter. An expected nul"lu" ,urrg" rpecific lilter is provided by the manufacturer. 'l'ho difl'ueion air flow or forward flow method, which is based on deterr lng the uir llow the wet firter at u ,p""ifi"d pressure 1". soz ot 'cross ttutrble-polnt presnurc). I[ is a more sopiristicatei method than bubl polnting, requlrlng n special test kit (available from manufactur.rry. an pected valuo rarrge, or built in pass/fail, is providcd by the ,unui""tu
60

'-

{C} 'ft,t* is a new technology and further developments are likely. {bl 'ttt,' cl'l'cct of filtration on the levels of important proteins (e.g. in the nterliurn or product), must be tested. [] flent,tuul of large (>80 nm) viruses, e.g. infectious bovine rhinotracheitis vhrrn (lllRV), Epstein-Barr virus, and retroviruses, is most effective.

(ltlrer mcdium-sized viruses (c. 50-60 nm), e.g. parainfluenza-3,

and

tl

:fltrrll viruses, e.g. picornaviruses, may also be removed to a lesser extent. E{fttei"rrey can be increased by using multiple units.

[1

Peter L. Roberts

2: Sterilization
(ME(') which provides protection to both the operator and the cells. Laminar fiB* cnbinets should only be used for preparing media and horizontal models FU:t rrol be used for handling cell cultures (see Chapter 4). Class I MSCs Bffer tro protection to the work but are recommended because they give a hlgh arrl consistent level of operator protection, when handling pathogens in hrccr.l group 3 (18). If it is considered essential that cell cultures and hazard provide 33up 3 agents are handled together in a Class II MSC, in order to pf6teetion to the cells, then a case must be made to the Health and Safety E*ecrrtivc or other relevant agency. Additional safety testing of the MSC is llkely t,r be required. Alternatively, a fully enclosed Class III cabinet would pFevlrlc lull operator protection and some level of protection to the work BEarr*c the air, although not laminar, is filtered. A Class II MSC is in fact leqrrrrlc lbr work with most common microbiological agents, i.e. hazard fFau;r ,2 pathogens (18). Guidance on the routine use and maintenance of Ablncts is given in Protocol 10.

(f) Units are designed for single use.

user, have been developed.

(e) Costs are high relative to standard filters.

(g) Methods for testing the integrity of filters, that can be carried out by t
-.virus-removing filters have found application, in the cell culture field, :]llg1l* viruses such;s_bovin_e viraf diarrhoea virus (BVDVj u'a ott larger viruses such as IBRV and parainfluenza-3 r.o- doui*'J"rrr-, manufacturer of bovine serum has used six filter units of 0.04 pm'to re and other possibre virar contaminants (15). virus ntte^ ar" curren PYDV i" rhe pharmaceutical indusrry for use with a range of c L:,:li ":llyl"d culture-derived or plasma-derived biological products and further advances this new technology are likely.

particulate air) filters are used to filter large volu of air in sterile or clean rooms, and in laminar flow or microbiol'ogical sa cabinets (L6,r7). Such filters act- as depth-type filters and are capable removing >99.97"/" of particres of 0.3 pm or-larger. They are thus effecti against not only bacteria but also viruses which exist in tne atmosphr attached to dust particles and riquid droprets. In the pharmaceuticar indus standards exist for the quality of air i.e. the maximurn level of farticles ar viable micro-organisms permitted for a particular crass of room or cabinet. F, asepticfilling, an environmentwithno mbre than l00particlesof >0.5 pm/ foot (3530/m3) is required to meet class M3-5 of ub rederal Standard 2 other clean room classification standards, e.g. BS5295, exist. The vari types of cabinet that use HEPA firters are lirt"d in Tabte 3. celr curt should be considered as a potential source of viruses (see crrapters 4 and and it is thus best to handre them in a crass II microbiorogicaisalty cabi
Table 3, Types of cabinet used for handling cell cultures Protection "

8.3 HEPA filters H"Efl (high. efficiency

Typc

Operator

Product

Air
H

lamlner flow cabinet (vertical)

Lamlnar llow cabinet (horizontal) Mlr;roblologlcal aafety cabinot (Class l) Mlcroblologktol rafety cobinet (Class ll). Mlerolrlologloal rdtoty cablnot (Class lll)

+
+
+

l
+
+
cabinet.

+
X

V,X
E,X

dPtottrrlhlr | I ), rro prolpolkrrr ( ), " vartlilel hilrrner 6rr rrow {V}, horrronrar raminar air frow (H}, air firtered on xhaust (X), or air on snlry and axheurt (E, X), uAlro oomnt'nry rlhttarr r. ec a crara I rominar frow hood or

|torformance testing llE|'n liltcrs and cabinets should be tested at least once a year, or every 6 fitaurllrs in the case of a room or cabinet used for the sterile filling of pharmacttlilrrl ;rroducts or for handling dangerous pathogens, to ensure they are perfurrrring correctly. Cabinet manufncturers and clean-room environment
i,5.
!t:t

52

Peter L. Robeds specialists are able to carry out the necessary tests. For MSC, these i challenging the HEPA filter with dioctyl-phthalate smoke particles of 0.3 pm diameter; and measuring the air in-flow and downflow velocity, necessary for the particular type of cabinet, to confirm they meet the requi standard, e.g. 855726 in the UK. An operator protection test, using iodide generated within the cabinet (Kl-Discus test), should also be carri out when the cabinet is first installed. This test is not strictly required unless the cabinet is moved or relatively high risk pathogens in hazard group (18) are handled. The level of particles within a sterile room can be monitored by the use of a particle counter. Additional biological tests product protection in rooms and cabinets used for sterile filling can be carri out by exposing bacteriological agar plates.

2: Sterilization
Pall Process Filtration Ltd, Ftltors, Pall Scientific and Technical Report STR1358. Pottarnouth. 5, 1. *{Jci;;" Laboratories (1987). Art to science in tissue culture, E' Spier and J' B' R' *iltprr, c. J. (1985). in I'itmat cell biotechnolga @d' blfntftg, Vol. 1, pp. l4l-f/'. Academic Press, London'

E;iii;r,

t.

H.

(;il

(1993). Laboratory-acquired infectiow, 3rd edn. Butter-

?0fthfi, London.

(1990). categoriz.atio.n of Patho' Atlvinory committee on Dangerous Pathogens. ind catego'iei of containrnent' 2nd edn' HMSO' ffflns uccording to hazard

Acknowledgements
I am grateful to the engineering and quality control staffat BPL for in on the routine maintenance, testing, and validation of sterilizers. I would like to thank Paul Harrison for useful discussions and Christine Thompson help in the preparation of the manuscript.

References
S. S. (ed.) (1983). Disinfection, sterilization and preservation, 3rd edn. and Febiger, Philadelphia. 2. Russell, A. D., Hugo, W. B., and Ayliffe, G. A. J. (ed.) (1992). Principles practice of disinfection, presewation and sterilization.2ndedn. Blackwell Sci

L. Block,

Publications, Oxford.

3. Threlfall, G. and Garland, S. G. (1985). In Animal cell biotechnology (ed. R. Spier and J. B. Griffiths), Vol. l, pp. 1?j-40. Academic Press, London. 4. Department of Health and Social Security (1983). Guide to good
manufacturing practice. HMSO, I-ondon.
5. Prusiner, S. B. (1991). Dev. Biol. Standard.,75,55. 6. Department of Health and Social Security (1980). Sterilisers. Health memorandum, no. 10. HMSO, London. 7. Bowie, J. H., Kelsey, J. C., and Thompson, G. R. (1963). Lancet,i,586 8. Brock, T. D. (1983). Membrane filtration. Springer-Verlag, Berlin. 9. Gelman Sciences (1993). Laboratory Solutions,l, L
10. 11.

Knight, D. E. (1990). Nature,343,2l8.

Dileo, A. J., Allegrezza, A. E., Jr and Builder,

182,

S. E.

(199). Biotechnology,

12.
13.

Dileo, A, J. and Allegrezza, A. 8., Jr (1991). Nature35l,420.

Manabe, S, (1992). ln Animal cell technology: basic and applied aspec* (ed. Murakami, S, Shlrahata, and H. Tachibana), pp. 15-30. Kluwer Academic llshoro, Dordrecht. 14. Pall Prococa Filtratlon Ltd, Rctcntion of Viral Contaminants by 0.04 pm Nylon 54

6[

Culture media
T. CARTWRIGHT
and

G. P. SHAH

Introduction
grow cells in vitro, culture conditions must mimic in vivo conditions with llltpect to temperature, oxygen and CO2 conceffiotr, pH, osmolality, and il[trition. Most basal cell culture media cannot, by themselves, support the of cells and it is common practice to supplement cell culture media animal sera. Growing cells in serum-free media has many advantages but ihe ideal general purpose serum-free medium has not yet been developed (fnd is almost certainly an unattainable goal). The main functions of cell ,fllture media are to maintain the pH and osmolality essential for cell viability |nd to provide the nutrients and energy needed for cell growth and multiplifftlon. The temperature and oxygen and CO2 content of the cell cultures lllUtlt also be controlled. A complete cell culture medium can be considered to composed of two distinct parts:

(E) a basal medium that satisfies all cellular requirements for nutrients

(b) a set of components that satisfy other types of cellular requirements and permit growth of cells in the basal medium
A nutrient is defined as a chemical substance that enters a cell and is used as Slthor a structural component, as a substrate for biosynthesis or energy ltetabolism, or in a catalytic role in such metabolism (1). Anything else heecled for cellular proliferation is normally classified as a supplement, inEluding all undefined additives such as serum and other biological fluids.

U, Basal media
culture medium is by far the most important single factor in culturing tnirnal cells. The function of this medium is to provide an environment for turvival and also to provide substances required by the cells which they gannot synthesize directly. The composition of early tissue culture media was bansd on biological fluids such as plasma, lymph, and serum, and tissue gxtrects especially of embryonic origin, Basal tissue culture media wcre
*l'he

T. Cortwright qnd G. P. Shoh

developed to include only the minimal components which were essential

3: Culture medio

growth.

important physiological roles Efreets, the inorganic ions used have other of membrane potential and as cofactors in enzyme iaeiUOing the maintenance

2.1 Types of basal medium

There are four main categories of basal media for mammalian cells several categories for insect cells. These are

o Eagle's medium and derivatives, e.g. BME (basal medium Eagle's EMEM (minimum essential medium with Earl's salts), AMEM (minimu essential medium with alpha modification), DMEM (Dulbecco's modi
Eagle's medium), GMEM (Glasgow modification of Eagle's medium),

are chiefly Eaett,roi and in cell attachment. The inorganic ions employed C**,Cl-, S@-, tcl31, and HCo3. When necessary, osmolie*, K', Mgt*, of NaCl. lglity rn,,y be adjusted by modifying the concentration =:'M1rrt RSsr oo not contain the nutrients required by cells for long'term included' The four main FElntenilnce or growth although glucose may be
legrrrics of BSS are

JMEM (minimum essential medium with Joklik's modification) o Media designed at Roswell Park Memorial Institute (RPMI), e.g. R L629, RPMI 1630, and RPMI 1640 o basal medium designed for use after serum supplementation, e.g. Fischer Liebovitz, Trowell, and Williams' o basal medium designed for serum-free formulations, e.g. CMRL 1 Ham's FLO and derivatives, TCl99 and derivatives, MCDB and deriva NCTC and Waymouth
For insect cell culture, the basal media (designed empirically) are

! Ecrle's balanced salt solution (EBSS) ! Bulhccco's phosphate-buffered saline (DPBS) ! Henk's balanced salt solution (HBSS) ! Eagle's spinner salt solution (ESSS)
equilibrated with air while EBSS and $Egli nrrct DPBS are intended for use with a gas phase containing 5"/" COz in order to ibBS r.quite equilibration
EBlntni,t the correct PH.

lrl,g lluffering

sYstems

o Grace's medium o Schneider's medium o Mitsuhashi and Maramorosch medium o IPL-41medium o Chiu and Black medium o D-22 medium
Most cell lines derived from cold-blooded vertebrates can be cultured one or more of the above basal media when supplemented with a biologi fluid. However, an allowance is usually made for differences in salt co tion to obtain the optimum osmolality. Different incubation temperatu may also require adjustments to be made to the composition of components, since pH may change with temperature due to alterations in solubility of CO2 and in ionization and pKu of buffers.

for evolution of co2 and Eni*or. ,u"oia .reeo to be buffered to compensate of glucose. Media have EE pt,nto"tion of lactic acid from the metabolism often at a final conHditl,,nutty been buffered with a bicarbonate buffer, forms a buffering system with dissolved *?lraf i.r,r tt Z+ -tr,t. Bicarbonate when cells are growing at low cell E, pr.r.tu.*d by growing cells. Horvever, CO2 may be produced to amaintain *Eift' or are i" a tag piur", insufficient cultures need to be grown in H r=q,,ir"a optimal lU. fo. this reason, these is both cheap and non-toxic to S=etrri,,rptrere of s-ioy. Co2. Bicarbonate buffering in the physiological E'==,'=tt* liut its pK" (6.1) resulis in sub-optimal HCO3 but high PO3o conieig=, S,,r" rn"Oiu are designed to coniain low

2.2 Constituents of basal media

'l'hc common constitution of basal media may be considered as follows.

2.2.1 llalnnced salt solution

Bulanced salt solutions (BSSs) have been used since the earliest attempts cell crrlture in vilnt. A llSS is composed of a combination of inorganic sal thnt rnuintain plrysiological pl{ ancl osnrotic pressurc. ln acldition to
58

in a CO2-enriched Fa?uti,,,.,', and therefor" do not require incubation as a buffer in some iiao*t,t,"r". sodium p_glycerophosphate is also used bicarbonate EFmuinri.rns. Each bisil *editr- has a recommended maintain correct pH and and $-He.errtr.lfion and CO2 tension to achieve gssrrlrrlity (Table I). =-F,,, ,,,,ir" effective buffering, without the need for elevated CO2 levels' a most widely used of these faaEe .,1 organic buffers "un b" employed. The acid)' b_uiF;,- is frepes (N-2-hydroxyethyipiperazine-N'-2-ethanesulphonic pH rangeT.2-7.6 and is more resistant F:epes i* ,, u"ry effective buffer in the with both F-iepirl pl{ ciranges than bicarbonate. Some media are buffered is both expensive and toxic to the E-i=*ri,,,,,u," and tlepes' However, Hepes preparations have ettr u, rrrncentrations above 1fi) mM. Impurities in Hepes at lower effective Hepes concenilac 1r."" rcported to cause cytotoxicity have been used are Fjiftu,,*, Other organic buffers, related to Hepes, which -E:i acidl and Bes 1,V rr'(hyclroiymethytmethyl)-2-aminoethanesulphonic
5g

T. Caftwright ond G. P. Shoh

Table 1. Recommended CO2 concentration (gas phase) to use with common basal media
Basal medium
NaHGO3

3: Culture medio
The importance of other water-soluble vitamins is less clear. becn reported to be essential for some cells and is included
Few data are available on the role of fat-soluble vitamins
109

Vitamin

B12 has

inFl2

medium.

in medium. Medium

concentration (mM) Eagle's MEM (Hank's saltsl

Grace's (Hank's salts) IPL-41 (Hank's salts) TC 100 (Hank's salts) Schneider's (Hank's salts) IMDM TC 199

% CO2 in gas phase

contains both vitamin

and vitamin E.

4
4 4

4 4
36 26 29 24 14 44

Atmospheric Atmospheric Atmospheric Atmospheric Atmospheric

5\

DMEM/Ham's F12
RPMI 1640

Ham's Fl2
DMEM

5 5 5 5 10

8,2.6 Hormones and growth factors Hormones and growth factors exhibit a variety of different effects on cells. These are included in some media (especially serum-free media) at relatively low concentrations. Insulin and hydrocortisone are main examples but growth faetors like NGF (nerve growth factor) and EGF (epidermal growth factor) havc also been used as well as certain interleukins, colony stimulating factors, Bnd fibroblast growth factors (FGFs) (for examples, see Chapter 6).

t,2.7 Proteins and peptides

Although an absolute requirement for proteins and/or peptides by cells in Cglture has not been established, relatively few media have been formulated ln wtrictr cells grow rapidly in the total absence of proteins or polypeptides. ommon examples of protein supplements used are fetuin, a-globulin, fibroncctin, albumin, and transferrin.

[N,N-bis-(2-hydroxyethyl)-2-aminoethanesulphonic acid]. Although

organic buffers function without COr, it should be remembered that bicarbon is essential to cells as a nutrient independently of its buffering role and suffici

bicarbonate for this requirement must always be present in the medi Actively growing cells will usually themselves produce sufficient CO2 for

t,2,8 Fatty acids and lipids

An with the proteins and peptides, there is no consensus regarding an essen-

2.2.3 Energy sources

Carbohydrates are a major energy source for cultured cells. Glucose is t most frequently used sugar. Other sugars, e.g. maltose, sucrose, galactose, and mannose, may also be included. Glutamine can also supply major proportion of the required energy in some cells.

tlsl role for lipids in cell culture. However, fatty acids and lipids are important omponents of several serum-free media.

t,2.9 Accessory factors

$nrongst these are the 'trace' elements, especially iron, zinc, copper, and

Glenium.

variety of other compounds, including nucleosides and tri-

2,2,4 Amino acids

Most animal cells have a requirement for the essential amino acids, i.e. which are not synthesized in the body. In the human, these are argini cystine, histidine, isoleucine, leucine, lysine, methionine, phenylalani threonine, tryptophan, and valine. Cysteine and tyrosine are also included tltis group to compensate for inadequate synthesis. Most animal cells huve a high requirement for glutamine. Glutamine acts both as an ene $ourcc nnd as a carbon source in the synthesis of nucleic acids. Other luci(1il ntc ol'ten rclded to compensate either for a particular cell type's inca Ity lo rnrke thenr or bccause they are made but lost into the medium.

Earboxylic acid cycle intermediates, may also be added to the medium.

1,2.10 Antibiotics
Although antibiotics are routinely used in laboratory-scale tissue culture, they lhould ideally be avoided since resistant micro-organisms may develop and Eell growth and function may also be adversely affected. Whenever possible, gnlibiotics are not used in industrial-scale cell culture where reliance is placed on ;rlant which is correctly designed and of appropriate quality to maintain pullure sterility. The use of antibiotics in biopharmaceutical production is not rendily acceptable from a regulatory vigwpoint. When antibiotics are to be used in cell culture, the key factors governing llrcil choice are

2.2.S Vltnmlnn
Severttl vltunlinB ol'llre I] group are necessary for cell growth and multipli tlon, Mntty vlturnlrrx llre precursors for cofactors. The vitamins most com udtled lrt bassl me(lln nre /ral(r-itmino benzoic acid, biotin, choline, folic nieotinlc ncid, panlotltenic ncitl, pyridoxal, riboflavin, thiamine, and i

I ubsence of cytotoxicity t broad anti-microbial spectrum I ucccptable cost t nrinimum tendency to induce formatitln tlf resistant micro-organisms
01

T. Cartwfight ond G. P. Shoh

Mixtures of penicillin (100 IU/ml) and streptomycin (50 pglml) are the frequently used anti-bacterial agents. Gentamycin (50 p,g/ml) is more expe sive but is widely used to treat persistent contaminations. Amphoterici B (2.5 pg/ml) is the most commonly used anti-fungal agent, but is cy for some insect cells. Nystatin (25 pglml) is also an effective anti-fungal in tissue culture medium. For further information on antibiotics, see Chapter
JUat

3: Culture medio
before use. Glutamine is a key metabolite for growth of animal cells but it relatively unstable and decomposes to produce ammonia which is toxic to colls. The dipeptide glycyl glutamine has been shown to be an adequate iubstitute for glutamine for some cell lines, and is more stable than glutamine durirrg both autoclaving and storage (2).
lg ttlso

2.3 Choice of basal medium

The choice of medium is not always obvious and frequently remains empi in spite of many years of exhaustive research into matching particular medi to specific cell types and culture conditions. Useful information is usu available in the literature or from the source of the cells. As a general gui BME will usually support the growth of continuous cell lines, e.g. HeLa, cells, BHK-21, and. primary cultures of human, rodent, and avian fibrobl RPMI medium is intended mainly for cultures of human haemopoietic while Fischer's medium isintendedprimarilyformouse leukaemiccells. Iscove modified Dulbecco's medium (IMDM) is widely considered best for cells of haemopoietic origin and supports the growth and differentiation both human and murine primary bone marrow cultures. Most insect cells, e. Sf9, Sf21, Bombyx mori, Trichoplusia ni, and Drosophilawill grow in Grace medium supplemented with fetal calf serum.

1,0.1 Equipment for preparation of media Whcther preparing medium from powder, concentrates or the constituent ehemicals, the following equipment is required:

I ltigh purity chemicals and biologicals I gttod analytical balance I hot plate with magnetic stirrer I vrtlumetric flasks of various volumes I pll meter
|
0ilnlometer

a Bterilizing equipment: autoclave and membrane filtration

Whcn preparing larger volumes of medium, it may be more practical for lEme purposes to measure weights rather than volumes. A large capacity bElarrce may also be appropriate in such cases.

2.4 Preparation of basal medium

Contamination of medium with micro-organisms, e.g. bacteria, 1leaSt, fungi, and with noxious chemical substances, e.g. traces of heavy metals, the greatest hazard in media preparation. For this reason, particular must be taken in the selection and preparation of materials. High purity should be used (see Chapter 1). Biochemicals should be of analytical Items of glassware used in dispensing and storage of reagents and media m be cleaned very carefully to prevent traces of toxic materials from contami ating the inner surfaces of vessels and thence becoming incorporated into medium. Basal media are frequently prepared by diluting a series of solutions, e.g. amino acid and vitamin concentrates, in water. These s solutions are stored separately in conditions appropriate to the indivi components. Incompatible substances are kept separate until they are m togother to make the complete medium. 'l'hc complete medium is usually sterilized by filtration (0.1*0.2 pm). mcdir can ulso be sterilized by autoclaving, e.g. EMEM, but care must tuken to sltbilizc thc B vitamins, and glutamine should be substituted glutunrute or udded al'tcr autoclaving. Powdered media are prepared
clissolving the powtler in the rccommended amount of water and ensuring

1,4.2 Preparation of media from powder FEr lirrge-scale tissue culture applications, it is often economical and practical tg nrrrke up single strength media from powder as outlined below. Protocol 1. Preparation of media from powder
Using a graduated container of appropriate capacity, dispense 90o/" al the required volume of tissue culture grade water, The temperature of the water should be 15-20'C.

Add the appropriate amount of powdered medium whilst gently stirring the water. Stir until all the powder has dissolved. Do not heat the
water, Rinse the powdered medium container with a small amount of tissue culture grade water and add to the bulk volume.

4, Add the required amount of buffer solution and any other additives.
E. For bicarbonate-buffered media, adjust the pH to 0.2-0.3 pH units below the desired pH using 1 M NaOH or 1 M HCl. Stir gently while adjusting the pH. The pH will normally rise by 0.1-0.3 pH units during filtration.

nll the conntiluentr ule completely dissolved. Unstable constituents

sodiunr lricartrorrale or u*corbntc) are usually added as a sterilc conce

(e

6" Make up to the final volume with tissue culture grade water and mix by
gontle stirring.

T. Cartwright ond G. P. Shoh

Protocof 1. Continued 7. Sterilize by filtration using a membrane with a pore size of 0.22 y,m or less. A positive pressure (3-15 p.s.i.) is recommended to minimize the loss of CO2. 8. Store at 2-8'C in the dark.
{a) runalysis as defined

3: Culture medio

lhc concentrations of key elements should be confirmed by chemical in the formulation (0 it should comply with standard protocols for sterility (:t lhc endotoxin level should be less than 1 nglml Tlto basal medium must also be able to support the growth of appropriate
@lh through at least two sub-culture generations (serum supplementation is Hfttttlly required to achieve this).

2.4.3 Preparation of single strength media from 10x

concentrate
To prepare single strength media from a 10x concentrate, the following need to be taken.

Protocol 2. Preparation of media from 10x concentrate

1. Using a graduated container of appropriate capacity, add the appropriate volume of the concentrate to 8O/o ol the required volume of tissue culture grade water and mix gently.

Storage of medium general rule, medium is best kept at 4'C and the shelf-life at this It tr tlttpcrature does not usually exceed 3 months unless otherwise specified by tt," manufacturer. Once glutamine has been added, the shelflife is, in lettcrirl, reduced to 2-3 weeks although individual media can last much Ittttgcr than this at 4'C or even at room temperature. Media which contain hhllc constituents should either be used within 2-3 weeks of preparation or

l,0,ll

lftttctl at -20'C.

2. Add the required amount of buffer. 3. Add the appropriate volume of 200 mM l-glutamine and any other
additives.

Serum F,l Why use serum?

Hlrlorically, the first tissue culture experiments were performed using animal bttly lluids such as lymph to support cell growth. When Eagle and others, in the lnte 1950s (3), produced basal media containing amino acids, carbohytlrrrtc, vitamins, and minerals, it became apparent that supplementation of ffictliurn with body fluids was still needed to provide unidentified, but essenliul, l'actors needed for efficient cell growth. Supplementation of basal ffirlirrrn with up to 20"/" of animal serum became widely used. Because of its flt,lt trrntent of growth factors and its low gamma-globulin content, fetal hrvirrc serum (FBS) has been adopted as the standard supplement. FBS is now nrost frequently used at 107" concentration although this may be Clttrrrgcd for specific applications. Advantages of serum use include the
fitllowing:

!,

4. Follow steps 5-8 of Protocol

1.

2.4.4 Precautions to be taken during preparation of media o avoid using partial quantities of prepackaged powder media because of
hygroscopic nature of many medium components

o prefilter water (0.1 pm) prior to o avoid excess acid/base additions

use

.
o

ensure that mixing vessels are properly cleaned (depyrogenation may be appropriate for some applications) perform filtration as soon as preparation of the medium is complete

2.4.5 Quality control of basal medium

Complete basal medium should satisfy the following criteria:

(a) it should be a clear solution (b) it should have the correct pH at room temperature: this is

specific

(c)
(d)

individual media osnrolulity should be correct-this is specific usuully in tlre rangc 280-300 mOsmol/kg
rts

to individual

media

juclgcd by 11l'1,() analysis, amino acids should be present at

tiulrs cunninlent with tlro l<lrmulation

of most of the factors required for cell ploliferation and maintenance. lh) Scrum is an almost universal growth supplement which is effective with rrrost cells. Using serum-supplemented medium therefore reduces the rrcccl to spend time developing a specific, optimized medium formulation lirl cvery cell type under investigation. Sclum buffers the cell culture system against a variety of perturbations Ir' I rrrrtl toxic effects such as pH change, or presence of heavy metal ions, prrltoolytic activity, or endotoxin, 'l'ltrst' points are discusscd in morc dctail in Scction 3.3.
(rt I

Strrum represents a cocktail

05

f
T. Coftwright ond G. P. Shoh
The use of serum also imposes a number of difficulties (discussed in Secti 3.4) which impact on the safety, reproducibility, and cost of biopharmace cals produced in animal cells. These difficulties can be minimized by ca selection and validation of serum sources. Although almost all new manuf turing processes using animal cells are designed for serum-free medium i order to avoid these difficulties, many existing processes still use FB supplemented medium. This situation is unlikely to change fundamentally i the near future since regulatory constraints generally make it impractical a uneconomic to alter existing processes.

3: Culture medio
ffiUlligrlicity of growth-promoting, cell protection, and nutritional factors that

It Crtrrltins. These can be divided into specific polypeptides which stimulate Fll growth (growth factors), carrier proteins, cell protective agents, cell IttAehnrent factors, and nutrients (some of which may be small molecules Thlclr rrre attached to carrier proteins). Some serum macromolecules can fill

Fttrc tlran one of these roles.

}fl,|

3.2 Types of serum

Despite its high cost, FBS remains the most frequently used serum for medi supplementation. Several different types of serum have been proposed cheaper alternatives to FBS. Calf serum is quite widely used industrially is available either as newborn calf serum (which has high levels of biotin) as mature calf serum. Newborn calf ^y-globulin levels are high as a result of ingestion of colostrum immediately after birth. Adult bovine serum is occasionally but is not usually as effective as FBS or calf serum. Horse se is also used, particularly with some human cell lines. The use of human se has been proposed for some fastidious human cell lines, but it is not c established that human serum performs better in general than FBS. Whereas FBS is usually collected at the abattoir, generally by cardiac puncture, calf serum and horse serum can be produced from ' animals. In the donor system, herds of virus-screened animals are maintai isolated from other stock and are used exclusively for serum production. health of each individual animal is constantly monitored, with special re ence to virus infection. Animals are bled at intervals by aseptic venepunct as for human blood donors. Advantages of this system over sla

Growth factors Folypr,ptide growth factors are of particular importance in serum. These l=t(l kl)a proteins act via specific cell surface receptors as signals which lllttttrlrrtc cell proliferation or differentiation. In some cases, the presence Ef eer.tain growth factors may not be stimulatory as such but may still be lllpttlirrl since deprivation of the factor initiates a pre-programmed autodrlructive sequence of events (apoptosis) which results in the death of cells EVetr lhough they may be fully provided with nutrients and be maintained
Efldet' optimal culture conditions (4).

llll'l'crcnt cell types have different growth factor requirements and the same llttwtlr l'actors may stimulate or inhibit depending on the cell type and the ifnwttr lactor concentration (5). Different types of serum (and different bllClr.* of the same serum type) may contain different absolute and relative bVelx ol different growth factors. This is one of the main reasons why growth lertlrrg of serum batches is necessary to ensure satisfactory performance with lhe rpccilic cell line of interest.

1S,2 Albumin
Allrrrrnirr is the major protein component of serum and exerts several effects lltlr,lr contribute to the growth and maintenance of cells in culture. It functions pt il r,rrlricr protein for a range of small molecules, particularly lipids. Transport ttf lirtty acids is an important function of albumin since these are essential for llr lrrrt are toxic in the unbound form and are also very poorly soluble in water. Elcr uitls and fat-soluble vitamins may also bind to albumin. (Other lipids such as holcstcrol, cholesterol esters, triglycerides, and phospholipids are transported

collection are:

o better control of animal husbandry o comprehensive knowledge of the animals' health status full control of blood collection and processing in a fully integrated
manufacturing practice (GMP) process improved consistency of serum quality since animals remain in the herd for several years o full traceability from bottled serum back to the individual animal if req
Serum from donor animals is available from a number of companies ing thc Salzman Corporation, Bocknek Ltd, and TCS Biologicals Ltd.

3.3 Constituents of serum

Serum is urr cl'l'cclivc growth-promoting supplement for practically all types cell (for $ontc cxr:el'rliolrs, scc Chaptcr 6) because <tf its complcxity and
00

ilt rcrrrrn in micellar form complexed with specific lipoproteins.) Allrrrrrrin also has specific binding sites for thyroxine and for metal ions such *. There is evidence that albumin may also bind other metals I Bt Nir and Cuz Fltrl rrlso carry other, unidentified components which support cells in culture. 'l'ltt. rrlrsorptive capacity of albumin also enables it to act in a detoxifying role hy lrirrtling toxic metal ions and other inhibitory factors. Allrrrrrtin also functions as a pH buffer and protects cells against damage by therrr lorccs that may occur in stirred or pumped culture systems. This latter nl'|elt irppcars to be entirely mechanical and related to the hydrodynamic Fl'npc!li(js of the medium since cells become protected immediately after rilltiorr ol'albumin, before the albumin prcparation has had time to exert aty possiblc mctabolic cffccts (6).
g7

T. Cartwright ond G. P. Shoh

3; Culture medio
Table 2. .Nutritional and protective factors which may be sup' plied by serum Factor Specific growth factors: EGF, PDGF, lGF, FGF, lL-1, lL-6. insulin
Trace elements

3.3.3 Transferrin

transferrin/Fe3+ complex is taken up via specific cell receptors and, a release ofthe iron, apotransferrin is liberated from the cell and recycled. It not clear if iron transport is the only role of transferrin; some reports sugg( that it may also transport vanadium, and others that it might have a wider in heavy metal detoxification. copper may also be transported by a carrier protein (ceruloplasmin) and chelated form by small peptides such as Gry-His-iys (GHL, liveigrowth factor

in medium. Transferrin (siderophilin) is the major iron transpor protein in vertebrates, representing 3-6"/" of total ."iur protein. ;
soluble

prgs,ented in an inappropriate form in the medium. Iron salts are also sparingl

lron is an essential trace element for cultured animal cells but can be toxic i

Goncentration 1-100 ng/ml

lron Tinc
Selenium (also Co, Cu, l, Mn, Mo. Cr, Ni, V. As,

1-10 pM 0.1-1 pM
0.01 p.M

Si, F, SN)
Lipids Cholesterol Linoleic acid

3.3.4 Anti-proteases
Serum contains two major classes of wide spectrum protease inhibitors, antitrypsin and a2 macroglobulin, each representing around 2/o of the tc serum proteins. Proteases are secreted naturally by many cell types (to extent which depends on culture conditions) and are ,n"d in the sub-cult of anchorage-dependent cells. The powerful anti-protease activity of seru prevents proteolytic damage to cells and to products.

c. 10 pM 0.01-0.1 pM

Steroids
Polyamines Putrescine

Ornithine Spermidine Attachment factors

Fibronectin

0.01-1 pM 0.01-1 pM 0.01-1 pM 1-10 pg/ml

Laminin
Fetuin

3.3.5 Attachment factors serum also provides attachment factors which facilitate the binding anchorage-dependent cells to the substratum. The major serum protein volved in attachment is fibronectin. Fetuin and laminin also play a role. A summary of the main components of serum that are known to important in culture medium is given in Table 2.

Mechanical protection Albumin Buffering capacity Albumin Neutralization of toxic factors Albumin Transport of metals
Transferri n/Fe3* Ceru loplasmin/Cu Protease inhibitors
2+

2-4 mglml
1.5-2.5 mg/ml 0.7-2.0 mg/ml

3.4 Potential problems with the use of serum

There are a number of serious disadvantages incurred when serum is used supplement culture medium. These have different impacts depending on different uses to which the cultured cels are put; in the proiuction of pharmaceuticals, compliance with rigorous rigulatory controls conce polcntial contamination by viruses and other adventitious agents is the pri concenl, whilc this may be of limited relevance in researclistudies. The elil'licultics e ncountcred when using serum are detailed below.

ct, antitrypsin ct2 macroglobulin

l,ur:k ol' reproducibility bnlclrcs viu'y considerably depending on the characteristics of Holrrce rrrrirnulr rnctl, orr.thc lbcd stuffs employed, on the time of year, llil'l'erenl bulclrr:s trorrruirr dil'l'crent absoluie and relative levels of grr l'ucloru, ('ertrrlrr lircrom nrrry bc cleficient in somc batchcs whilc others may prcsclrl rrl ereernivc, ilrlrihitory levcls lirr somc cell typcs.

ll.4.l

selrln

Vuri:rtions in performance of this sort are not tolerable in manufacturing reservation system where batches fft'r!('..sscs and are countered by the batch Ale lrr.ltl on reserve by the serum producer while the would-be user completes lerlirrg ol'samples for efficacy in his own specific system. The situation is also llrprovcd if the reserved batches are large enough to permit production over a t'orrsitlcrable period. trr r.xpcrimental cell biology, it is also important to be aware of the inherent Vntillrility ol serum which renders it very difficult to study the specific effects rtf lrrolcculcs such as growth factors, cytokines, adhesion molecules, or matrix Fuurlx)ncnl"s, all ol'which arc prosent itt unclcfined levels in serum. 00

68

T. Cartwright ond G. P. Shah

The presence of specific antibodies in serum may also profoundly affect results obtained. This is especially true in the culture of viruses. Antibodies

3: Culture medio
OF ANALYS$ FOETAL CALF SER(]M - CERTIIICATE

serum may result from a natural infection with the virus in question or related species (which may be transmitted transplacentally in some cases), from prior vaccination of the animals used. It should be remembered some antibodies traverse the placenta and that even FBS may contain si ficant inhibitory activity to infectious agents to which the mother had
exposed.

Serum may also vary depending on the quality and the reproducibility the procedures used for its collection. For instance, the length of time tween collection of the blood and removal of the cells is critical, and m be minimized if lysis of cells and the release of cellular contents (possi including viruses) is to be kept low. Sterility of the operation and seve other process parameters are also critical. Reputable serum suppliers invested heavily in the quality of their operation and serum producti GMP standards is a more complex process than many users realize. Reli tissue culture quality serum is accordingly an expensive product.

Bovine Viral Dianhoea Psrainfluenza 3 lnfectious Bovine Rhinohacheitis

a't37"e
mOsmolftg mg/ml
rng/ml

3.4.2 Risk of contamination

Serum can represent a major route for the introduction into cell cultures adventitious agents including bacteria, fungi, mycoplasma, and viruses. could be disruptive in research projects and dangerous in pharmaceuti manufacture. In order to minimize risks of contamination, suppliers apply rigorous health checks to the animals used, use GMP facilities collection and processing of serum, employ thorough quality control and ensure rigorous batch documentation to permit verification of the process. The process employs aseptic collection by cardiac puncture or venepuncture, aseptic clotting and clot removal, clarification, and sterili tion by filtration terminating in double 0.1 pm filters followed by filling. Several companies are now producing 40 nm pore size filters should provide additional safeguards by achieving better clearance of vi However, it should be recognized that some of the higher molecular proteins (e.g. IgM) may also be removed by such filters. Serum is then

ng/ml Pgl^l

-g/nl
EU/ml

Frimn Epittt"tiel Cell Growth Capacity capacity fiyel,tmritlyttidona Growth ldrtivo Cloning EfficiencY
yftrtoxicity check

lehllvo Plating EfficiencY

% of control % of control % of control % of control % of control % of control

Ereuttmntttkrn APProval

Signed

.'...............""""'Date .'..'...'..........."""Date

for microbial sterility, for

contamination by certain viruses, and for capacity to support the growth of test cells. An example of the quality tests routinely performed on serum batches is shown in Figure 1. As the recent outbreak of bovine spongiform encephalopathy (BSE) the UK illustrates all too clearly, this approach to testing cannot eliminate risks of contamination. As a further line of defence, regulatory a now specify thut tlnly serum from specified countries of origin, where lar agentr uf coneern nre thought not to occur, can be used in the of pharmaceulicnl, veterinary, and sometimes diagnostic products. restrictionn arb unlikely to nffect the research user, but are mandatory
70

idrl'l

l{olcase

Signed

F$lr t,
I

I t;BHil,J'i#:'J;r;" ,;;;;'Jrpprlirn" ::^ :'": documentation used to senerate ;'; this lppl er I o batc es ot "" -'-ro El* ::^t:originar ::i*::l::' :: : ::n""'4
t

regime which should be An example of the type of quality control testing

h

l,ffinrary .hoet Bhould o"

specification for examinition on request. The precise All type of serum in question'"ff sera depend on the specific lllgft l|n tho analysis certificate no d"t""trbl" conteminaiion in the sterility tests' (NB: for h$Ed rhould, of course, rnot" wlll be based on other virus tvpes') 5lrsnr ol non-bovine .tis;;:;1;;;;;;tiil

rr"if"[i"

71

T. Cortwright ond G. p. Shoh

3: Culture medio ) is such an expensive commodity there is a temptation for unscrupulous s to misrepresent the origin and the quality of the material (7). It is pgrtrnt, therefore, foruserswho require highquality serum of definedorigin Ugc u reputable supplier who has direct sources in the required production . Such suppliers will be pleased to provide comprehensive and verifiable, documentation for every serum batch and to allow audits of all stages the production process. This approach is important, both for validation of qunlity of individual batches and for continuity of supply. ontinuity of supply and consistency of quality are, of course, also importtrt the research user who may not be operating under strict regulatory rnints. Again this is best achieved by dealing with reputable suppliers Ituvo the physical and logistical resources to provide this service. It nlrould be noted that, independently of the regulatory needs of the tology industry, movement of serum between countries is also subject frtriction because of concerns by governments over animal health.

?"i::,::l:,,T* f can be tIii"o n q ^p--p-pioru",o,," or by #:::-::::*:::ii ?q:",: *:h-56'i f;t-1';t;;'-"",t"1 gamma-inadiati Heat inactivation (usuatv at 5ir;ffiTil:::l general, a' of these processes
t
.

of serum and because of this, an extensive ,black market, has sror up supplying relarively low price serum
of Joubtfur orilin g tle re quired ri gorous" staro".o,' f zl : risk of To reduce further the ":_.,rg virar conlminatron, serum also resurt

of biopharmaceuticals. This has resulted in the creation lif AT":?,:1""-"^l::":l^?l]:countries*iin-N"*z"uruna,-au,iJiu,a,,d f o ccupyin g th e his.!e1 uS rr ac::i N;,;,e";i;ffi i. ;.frJillllil;lLirl the cost

manufacturers

ffi;,l#;fi;

capacity and increased cost.

in

decreased growth-promot

3.4.3 Availability and cost

eesr;u; ffi;#H; il ;;";; ;i' o,l *in,,, significant investment and.operating i::*:j "',tl':::,: ^":.iT:111 : :"'v incurred b"";t;;'ft::,; "oi, -" *:l':i:"::,,Tyl::T:or,ig tg cvrl f.in"ipr".. correctry couected, p cessed, and validated FBS will always th;r"fr;;';;;;,"br;;i""ffiiil: of a tissue
rr ig_

Serum is a by-product of the meat industry. As such its supply (pa in the d iffe re nt p ::?,," _lr"".Tt supprv of geniine countries. The_ *'n;T.l,^""i Tll,:,ir--R-Jh", "r N"* y'Juna

h ;'ice.

Selection ofserum
ion of serum type and of serum batch is essentially based on empirical

ion by the user. When a given cell type is first grown,

it

may be

culture process.

3.4.4 Influence on downstream processing

l*:::::,1.::-:f^':r_1T I, when purifving a protein

-.

serum-containing medium because

i

-'--r e'^"vsr! rv r,ur'J [i;-3'-*l*,':':"-::::"1'"9u"i{*.;;i#;;iffi ',"l""}u.rv rr of ttre uantities i _, c"'', ;i Jilg" il**:t"1 * Iproduction ot p,ir" prot;i":,;*rJ.ii# "our tn c.mmerciat "li,'",li;
q

about^4-8 mg of protein io ,r," medium wh'e recombir ;uumDln proteins are frequently expressed at levers of tens of micrograms per mr. otthe required ::: :*y::,:1"^,",1r_lurificarionbe i-por;;i;';; protein may be diffic and in some cases it may even acceptable purification process. Monoclonal antibodies may be secreted by hybridomas at higher

tissue culture medium presenrs particular difficur secreted bv rh"

!njt:

d ;r ";il.;;;;#ffi;lion,.J

hwhile to test its capacity for growth on cheaper serum types, possibly rrg mixtures of different types of serum. Altllrugh serum offered commercially will have been subjected to thorough , irrcluding tests for cell growth performance (Figure /), it is still routine marry users to 'batch test' serum in their own system before buying. For llelum suppliers will generally provide samples free of charge from tlilTerent batches for the user to test in his own laboratory to select the

Iiffi'#%ff#;

"

n- (si gn

pro""r, cost and may determine :::.,1l':r:l]Ivcr ,f the crrtire pr.cess. In such d.;;;;;;;;;,i#:",},[I;: the via procdB' ,fi ile supprementation is a criticar disadvantag". - -"'
^rm

80%, of rhe totat

pro""rring

For rosron;.alreatry trincunsctr, *rr some apprications the geographic source $crunr may bo trerermrncd rry regurut.ry agencies. Becau-"se Gnim
72

3.tl Sourclng and solection of serum

l"speciar

rnost suitable for his specific requirements. The quantity of serum nliully required by the user is held on reserve until testing is completed. lry the user is primarily to check performance (Protocol3) but may lrrvolve verification of the absence of adventitious agents or analysis of $16 ffiy ltirrirrneter that is of particular importance for the user. It la irrrportant to bear in mind that tests of serum batches for growth or pforlrrctivity should always be performed in an identical system (in terms of . tA:ut rrrcclium, cell conditions, and culture configuration) to that in which the tsfurrr will finally be used, and should always include a previously evaluated tsfeterrcc scrum as control. It is recommended that the cells are sub-cultured Bl lesxt llrrcc times as indicated below. At each passage, growth results should FE lrrrurrrrlizcd relative to the reference serum. Yields within l-20yo of the tsfetrrrcc ricrum would normally be considered satisfactory. In addition to cell IlElrl, r'clls should also be examined at each passage for any indication of pftoluxicily or of abnormal morphology. l'ests of cell growth from very low tsedirrg lcvels may bc important lbr spccific applications. Tests of cloning Effielerrcy or of plating efficiency ntay be pcrl'ormed to assqss this.
78

T. Cortwright ond G. P. Shoh

3: Culture medio
have Erowins mammalian cells. Based on this information' many attempts

Protocol 3. Functional testing of serum batches The following is a test of the capacity of serum to support growth of the
required cells over several passages.

6a.n *""4"

1. Prepare flasks containing an appropriate growth medium supplemented with either 5% of the serum to be tested or 5Y" of a previously tested reference serum (use of 5% serum gives a more sensitive indication of the growth-promoting capacity of a batch than does use of higher percentages). 2. Seed the flasks with a number of cells appropriate to the cell line used (generally 1000-1500 cells/ml for attached cells). 3. lncubate the cells for 5-7 days, harvest the cells, and count. 4. Split the cells 1 :5 (or at other appropriate ratio); seed cells from the test serum flasks into fresh medium supplemented with 5olo test serum, and the reference serum cells into 5olo reference serum. 5. lncubate flasks for 5-7 days as before, harvest, and count. 6. Repeat steps 4 and 5. 7. Calculate growth-promoting activity relative to the reference serum (generally accept reference serum value 1207o).

replace serum in part or in full by serum-derived factors or by to reduce the serum require$mpt"tety synthetic media. one approach is culture medium with processed serum products' Coniiini Uy supitementing (CPSR) are prepared by processes that tfoltcct process serum ieplacements A"nn"a products with much higher batch to batch consistency than

i"

}iEi,f

Ginl

replaced Srntcin and endotoxin levels than serum. Natural serum can also be b:, gy ;;;pi";"nted/fortifi ed serum' Serum
.

cpSR products are derived from bovine plasma and have lower

and trace Ciu*rii'factors, hormones, proteins, other protein stabilizers, A;;;;;. S-;ch fortified ,"*- can be used at a much lower concentration
$en normal serum.

-1-

l"^tj*,*:t I j"?llj;

whenever possible

it is desirable properly designed serum-free

3.6 Serum storage and use

Serum is rapidly frozen by the supplier immediately after bottling and is at -20"C. Few data are available on the shelf-life of serum held in this but 2 years has become accepted as a rule of thumb. Measurement of stability in real time is difficult (due to the difficulty of standardizing

I I I

It rcproducible market h not reliant on the economics of the world cattle

rlrnplifies downstream purification inhibitors hns rro unknown factors e'g' viruses, or growth

serum-free cell A rrumber of cell types have been grown successfully in for one cell line. iturc media, usually in a medium spLcifically developed lines diffei greatly and success of a serum-free ffi=

culture over the period required) and time-consuming, while accelera degradation tests are not interpretable when applied to frozen However, the one rigorous real-time study of which we are aware that 2 years at -20'C is a conservative estimate of shelf-life and that could be extended to 5 years. When required, serum should be thawed rapidly and with gentle mixing minimize protein denaturation due to salt concentration effects. An wuter bath at 37'C is best although serum should be removed as soon as lhnwed and not allowed to warm up.
'l'hnwed $crum should be clear and there should be no significant tlttn, Once tlurwcd, scrum can be held at 4"C for a maximum of 2-3 Scrum rhorrltl ttol be re-frozen.

'"q"r..-L;;t;icell et-Elt,,"i formulation with one cell line i"tpit"n1"11ut,t:::11^:^t::t: gone inro developing $,i=g6;i;t;imitu, "ett lines. A great deal of effort has limited. However' with success has been EfUtr,-1r"" .edia, but until recenlly, of essential gro*ih factors and nutrients required by differFJlit",,tin"ution

*":

4.

Replruconlont of serum in medium

Muclt of the prorcrrt urrderstnnding of the nutritional requirements of cells culture atcms l'r'unr Llugle's work (3) on the fundamentul requirements

media have been formulated' Enl .*ffr, several very effect--ive serum-free =frOfr t lists the estabiished cell lines which are able to proliferate in defined a period of-'weaning off' serum Rgtllr,,,t without adaptation; those requiring and species are represented; iiE-rut, included. A-variety of tissue so^t'rces gl*ri' ,fr"r" are unique combinations of growth factors and hormones that 6)' The !f'nr,,r., .ptimal proliferation of specific "ill typ". (see also Chapter the polypeptide hormone' for Hu-*i .,t"ti*,ent iequirement app;ars to be ,,,rJ fo, the iron-transport prote.in, transferrin. other supplements iHliiir'i, growth hormones, polypeptide growth faciHeh,,t.' polypeptide and steroii albumin comkfr, 1,,,." clements, reducing agents, diamines, vitamins' and consideration for some bb*u,t with unsaturated fatiy icids. An important Hnll,rrti'ns is that animal-derived supplcments or proteins can pose conHiiiii,,,i,tt risks similar to those of sorum (see Section 3.4'2)media are now ==Buu,.r,,l commercially procluced, reudy to usc serum-free
76

3: Culture medio
o o

c o E E
o

6 F=E **69;g
@Oee

orN('tsrrl)(oFo)oFN FFFFNNN
c '6
o

NNNNNNN

cttrnGlFCoo)

E;E 5E

i;+fi 9J6iF

|Yalluble which have been designed for particular cell types. These are sumGltlzcd in Table 4. It should, however, be remembered that different strains !f thc same cell type may have different medium requirements, and that 'fine Hnlng' of these commercial media may be necessary to obtain optimum tslUlts with a specific individual strain or construct.

gE eePseEeEeEEeeee ieeEeie Fi ooo3000 a7 cccccFcitrcrtrq:cccc i*;*; oE 00000-oo.o0.o-ooooo cLccrEco.c : X P-^.8

I

*!i i;,=

-cL EO 5E

.=

E 6 6E O rD O O

.=

.E.g

.s @

q,

.E cr"H

.E

E O

o
.E

! --.!J o o
E

;sH$ @?oa= *

Irt

Design of serum-free media dclined serum-free medium is one in which a group of components

are rlated together to optimize performance of a single cell type. Each comincluded is of known purity and is present at a known concentration.

..r ii: Yu

FHirifisEFiH$ii rrj-'.-.rrjri jriu:u:": f; ririrrjtrj t'lj'ddtiriririH ii u #;9EHifisdFi. E ; rr-)C,Et4 IE t

EEEE E E &,E E E
rE

;I.gd. d,,,qfi s FeEdSI.giqad iSpp- ip E ;li5.

E

ar< =C

L (9

;:qt:
sba
>c

efig l
5 0P

important factors must be considered to achieve this goal. Amongst rurc the origin of the cell line, i.e. species and tissue, the compatibility of components and their interactions, and the specific application for the cell line is being cultured, e.g. production of biomass or generation pnrtluct. The two main approaches generally followed in designing a serumtttcdium are:
Rcduced serum: in this approach, the concentration of serum in the basal nrctlium is progressively reduced whilst other components, e.g. growth

5d PEE EI;:".

F, E:

F'e-n-D-'e-'e--e-|2-E-D-D-D-"e_D_'e_E;glEE_D_r_

FPFE sHEH

5 !,
E

cr
o o o o

E o

oss dtt E atgg ur!uJ LUlIJulullll>u,uJ oosul=uJr! > > ur > > o () f f > > ::HgE. SSS>>>S> l! L r O O cl L O = O O O > O O > > E O c O O u-'c to -' i; O= o-.t =

,:ss ssssS&:s a!e ::d t+t tttt:tt

E=

_o t3 d-

!'i E t''^ Eg:I

E

fuctors and hormones, are added to identify the factor(s) capable of tcstoring growth to the level obtained in the presence of serum. This Ittoccss can be very lengthy because at each change, growth assays using tlrc scrum-supplemented control and repeat verification assays need to be
rkrnc (30).
Busal medium: a different approach is to add components (singly or in combinations) to a basal medium in a stepwise manner until a medium is progressively 'built up' to give a similar or equivalent cell growth to the tc nr m-supplemented medium.

_9

E
E

Ei E Effi
3 3 -.E B E B:3:3E E 3 3 E E EE E E E E E g.EE E E;

EE58

':E
frfr=
E

.-; o E t: .8"34 b
E Ai I'iE
f c .: ,.'Q)

E:EET EgHE

approaches, the following critical factors need to be in designing an efficient, defined serum-free medium. *inrltlcrcd

frf

ultlt"t of these

3E4flE 'i ba'ab* 3+5n'E

Irl,l

.: E'

o e--

ti
!

sg

Fgu*Ea ffginii s* s a; s B,igE E r s s s s s aE

FFFFFE

f;E5i t'69ct
E tr.:;
.P

llusal medium lfte rclcction of basal medium can be extremely important in terms of energy lUtccs, buffers, and inorganic ions. Generally the starting basal medium kfntrrlrrlion is chosen on the basis of the known preferences of the required tsll llrr.'.

E:

E.E

,x

'aor9g

EEEEE Z.iiiEiSE E ;';sEP ?E;gs

lrl,t
frerc

l,iPids

FlI

e.E.i'r;t

l!
F(,

.i tl

rlfisrr*,Hfrgl*Esi igssEif

HiI.E ! I cl'o:
{E=8$
h

irrclude ethanolamine, phosphoethanolamine, sterols, fatty acids, and Fhugrlrolipids. In serum-supplemented media, they are usually carried on hgut'orrr,rlccules, principally proteins. In serum-free media, fatty acids are BtUulty provided in a bound form (eithcr to albumin or to other serum p1tlslrrr) rtr in the form of phospholipid'enclosed vesicles (i.e. liposomcs). If
77

3: Culture medio

3: rE
!J6 =@l

oqa

6:

og6: (rf,tttttd.hr

EE
o o o o o o
N
I

Y5 -- g

tsfum albumin is used directly as a lipid source, it should be noted that the lipid content of albumin may be dependent on the methods used Its purification; the solvent precipitation frequently used may result in stripping of lipid from the protein. It should also be noted that ized human albumin will have been stabilized with octanoic acid or hydrophobic stabilizing agents prior to heating and that it may be nt to replace these with more physiologically relevant lipids before ln cell culture. Recent developments have permitted the use of totally ic hydrophilic carriers such as cyclodextrins for the transport of lipids

6= =o^ o = !

EF g
^ Y 2

9
^.7 lrc 3'6
t@< I <

Buffering
maintain a proper environment

for the metabolism, growth,

and

9 < ti u
oitt

.Ei
9
a
o o

.9x

rning of cells. Major ions (Na+, K+, HCOI, and HPO!-) are usually cd as the principal components in pH control, along with H* and OH-, cnter into the ion balance. Other components, including amino acids, if

9 E

o5 nt=o

.g o

3
)

! e , 6 8,,i

Exo EE o@=

O 8 E ! ?

'i

C)

o bo ot

,r

o o
-9

'6
F

ti ,
o o

I FYH 9-q;=
Fo

JSPL

Eag

6 o
o I
O

in high concentrations, can contribute to the buffering power of a um, Besides bicarbonate, zwitterionic organic buffers like Hepes, Bes, 'lbs may be used in systems in which strict control of the gas phase is not rcd. However, careful consideration must be given to the concentration lltcsc buffers which can be toxic to the cells (32). Some of these buffers biologically important cations (33). A useful buffer for use in the rcc of low or no bicarbonate is sodium glycerophosphate (34). Trace elements rr.jor ions, i.e. Na+, K+, Ca2+, Mgt*, Cl-, HPO?-, and HCO', are involved in maintaining electrolyte balance and contributing to ic cquilibrium of the system. Trace elements are also included in many l'rce media because of their beneficial effects. Inter-relationships exist rr Fe2+, Znz+,and Cu2* ions which are needed for many cells. Most l'roe Co'* SeOS-. liltr-l'roe media also include Co2+ and SeO3-. Cells derived from heart and Itry tissue have a high requirement for K+ whilst Caz* is required for EEfllrol ol'mitosis (35) and the Ca2*/Mg2+ ratio is important in controlling cell pfttllli'rntion and transformation (36). Selenium is proving to be important for flHlty cc'll types (1). Other trace elements include Sn, V, Al, and As (37). Iron ll lterpucntly added as a transferrin complex but can also be added in other fuflttx srrch as ferric citrate, ferrous nitrate, or ferrous sulphate.

s2

t E,;
=6 ssi;.[;

o ot

.> c o
5 o
.9

!o
c)

-99 FE !3d g (,,q=

g:
EO

;60
EE O O^

E
3
6 J g 3

c o o

e i o

E
E $i

.J

TP8 P EHEas:HH*; o i E niPFEE:5==;; oo oJ ; o o oJ

c
a,

o o
C'

o
F

r
F

{ fi s I FE; F p f ir# , |
g"

i ! E

f =5:s*SEE;;Eg E3 dcE
.c

oa-

.9

t,
E

'.9

E gs b>F S;g P

ED

.s
c)

l,l,F

Mechanical stabilizers and adhesion factons

gr
5

c
o

o o o

&

fttr optirnal growth, cells grown in suspension culture require protection from lheur .lrrc to agitation (air bubbles, stirrer, and shaker). Shear damage can be Etlttccrl by increasing the viscosity of the medium. Carboxymethyl cellulose titl glrlyvinylpyrrolidone have been used lbr this purpose. The most widely Blerl nlrcnr protectant is Pluronic F-68. 'Ihis is a non-ionic block copolymer

T. Cartwright and G. P. Shah

with an average molecular weight of 8400 Da, consisting of a central block polypropylene (20"/" by weight) and blocks of polyoxyethylene at both Pluronic F-68 has been demonstrated to have a significant effect in protecti animal cells grown in suspension in sparged or stirred bioreactors. The tective effect is thought to be exerted through the formation of an i structure of adsorbed molecules on the cell surface. It is thought that t hydrophobic portion of the molecule interacts with the cell membrane, the polyoxyethylene oxygen may form hydrogen bonds with water molecu to generate a hydration sheath, which provides the protection from lam shear'stress and cell-bubble interactions (38,39). Cell attachment and growth of anchorage-dependent cells can be i by pretreatment of the substrate in a variety of ways. The substrate can treated with adhesive glycoproteins such as fibronectin, laminin, chondroit epibolin, or serum spreading factor (see also Chapter 4, Section 5.4).

3: Culture medio

!,!,8

Adaptation of cells to low serum/serum-free medium

fFhe following

protocol describes a generalized procedure to adapt cells to low supplemented medium or serum-free mediuln. The procedure should !funr Figurc that cell viability and protein synthesis are not compromised at any
Edaptation stage.

Protocol 4. Adaptation of cells to low serum/serum-free medium

Determine the optimal seeding density in serum-supplemented medium which allows for 5- to 15-fold growth during the experimont/ culture period. Using the above optimal starting density, replace the basal medium with the serum-free medium to be tested. Set up a series of cultures in this medium with varying concentrations of serum (e.g. 1-10%). Ensure that enough replicate cultures are set up to allow meaningful interpretation of results. lnclude selective agents to maintain gene copy number if recombinant cell lines are used. Select the culture condition that gives 60-80% (or more) of the cell growth of the control cultures and, using this condition, expand the cells to a larger scale (e.9. 24-well plate to 10 ml suspension to 50 ml suspension to 100 ml suspension, or 24-well plate to 25 cm2 flask to 75 cm2 flask to 175 cm2 flask). Check for expression of the desired characteristic or the correctly processed product. lf none of the conditions used achieve 60-80% of control growth, consider alternative serumfree media or adding supplements. Make afrozen bankof cells atthisstage (e.g. 5 x 106to 1 x l0Tcellslml) from an exponentially growing culture (see Chapter 4, Section g). Set up a second series of cultures as in step 2 above, reducing the serum supplementation further (e.9. 5% going down to 0.1%). Grow
cells for 3-6 days.

4.2.6 Selection of components

The following checklist is a useful starting point for consideration of
ponents for inclusion in a serum-free medium formulation:

o transport proteins, e.g. transferrin, bovine serum albumin, and o stabilizing proteins, e.g. aprotinin, bovine serum albumin, fetuin,
soyabean trypsin inhibitor

o growth regulators, e.g. insulin, hydrocortisone, and triiodothyronine o growth factors, e.g. EGF, FGF, NGF, platelet derived growth factor, insulinlike growth factors (somatomedins) o attachment proteins, e.g. fibronectin, collagen, laminin, fetuin, and
spreading factor

o crude extracts,
tissue digests

e.g. bovine pituitary extract, brain extract, liver extract,

o essential nutrients, e.g. cholesterol, linoleic acid, ethanolamine, and

elements

t,

4,2,7 Practical hints on solubilizing specific components o Riboflavin, folic acid, tyrosine, and cystine need dissolving in NaOH. o Insulin needs to be dissolved in HCl. r Fatty acids, lipids, and fat-soluble vitamins can be dissolved
solutions or attached to either protein carriers, e.g. BSA and or surfactunts, e.g. Tween 80 and Pluronic F-68. Pluronie F-68 is nrore soluble in cold water and, when making Pluronic solutiona, it chould be added to the water rather than the other way rou

Repeat step 3 until serum supplementation is eliminated. 7, Prepare a frozen bank of the serum-free adapted cells and check for the absence of mycoplasma (see Chapter 8, Section 4,3).

4'tr,11 Difficulties that may be encountered with serum-free medium Wlrcrr cclls are grown in serum-free conditions, they no longer benefit from the rrrrrltiple protective and nutritional effects that serum provides (Table 2). The robustness of the process in serum-free medium depends on attention to the lirllowing points:

Hypoxenthlne dicsolvec etsily on heating,

80

(e) ('clls appear more fastidious in the absence of serum: design of a dedicnted medium for each ccll typc is unually necessary for optimal results.
B1

T. Cartwright ond G. P. Shoh

(b) Culture conditions become more critical in serum-free medium: bet control of key process parameters (pH, oxygenation, etc.) is there
necessary.

3: Culture medio
cnabled the rapid measurement of low molecular weight nutrient utilization during cell culture. Medium optimization can proceed based on these data by

(c)

Serum-free medium has a reduced capacity to inactivate or absorb materials (e.9. heavy metals, endotoxin, etc.). Greater attention to purity of components and to depyrogenation is required. Antibiotics exhibit increased cytotoxicity in serum-free medium.
tu

(d) Specific shear protective agents may need to be added. (e) A significant adaptation period may be required before cells are
weaned on to serum-free medium (30). This makes the design and of serum-free medium a long and labour-intensive process.

tupplementation of rapidly utilized components and reduction of under-used eomponents. However, interpretation of such studies is complicated by the dynamic interaction between utilization of different components. This has led tcveral groups to base optimization studies on factorial experimental design tlmed at accommodating complex systems involving multiple interacting factors (41). Additional complications arise since cellular metabolism may alter during the course of a fermentation and different nutrients may become

eritical at different phases of the culture (42). T\erefore, reductionist epproaches to medium design should be applied with caution.

4,2,10 Direct influences of culture conditions and medium components on protein expression
Regardless of whether the medium used contains serum or not, many in the culture environment affect the quantity and the authenticity of protei

0. Influence of cell culture systems on choice of

medium
'l'hc efficiency of an animal cell culture system depends on the interaction of mnny different factors. The support that the medium provides for the cells is a erucial element but this in turn is influenced by the type of culture system in which the cells are propagated. C'ultures may be operated in a simple batch mode, or as fed batches, or as a pcrfused system. The cells may be grown in suspension culture or attached to turl'aces, or they may be grown at very high density (in excess of 108 cells/ml ln various types of plug flow reactor) or at densities around 105-106 cels/ml in
Flrrrple cultuies.

produccd by the cells. Apart from dependence on the basic capacity of t medlum to supply cells with nutrients and oxygen and to maintain osmolnlity, etc., protein expression may be affected by other, more su faetors. []or example extracellular matrices (ECM) can affect the regukrtion und maintenance of normal cellular functions (e.g. normal m mnry epithclial cells when cultured on collagen substrate produce 4- to
more casein than when grown on polystyrene substrate). Also, the tion ol' nutrient medium can affect the glycosylation of the expressed protein Clucose limitation often results in incomplete and/or aberrant protein gl sylution; ammonium ion accumulation in cell cultures can result in gl

protoins deficient in terminal sialylation; treatment of cells with differe hormones, vitamins, differentiation factors, etc., may often result in alte glycosylation patterns in glycoproteins secreted by the cells (40).

Cells may be enclosed in compartments which favour the development of krcal micro-environments (as in hollow-fibre bioreactors or macroporous eurrier cultures) or they may be fully exposed to the bulk medium. They may crpcrience significant shear forces (either by pumped medium flow or by ngitation) or they may be completely isolated from shear force (e.g. in the

4,2,71 Component interaction and factorial experimental

deeign

llrlcrior of macroporous carriers or when encapsulated). ln these different situations, the environment experienced by the cells nrudulates the support that they require from the culture medium.

ln lhc pnst, the design of serum-free media was predominantly empirical.

mttnl cttm nton approach was to reduce the serum supplementation progressi

d.l

Batch or perfusion cultures

('wetrtirrg ul'l'') and to determine the critical factors involved in cell

and pnrlcin cxpression.

Bntcrl olt lltis upproach, a database is now becoming established for nutrlllolml requlrenrents o[ the most commonly used cell types (see 6) and lltln lttformulion cnn often be applied generically. In general, t lormctl ccll llnen huvc u lower requirement for growth factors than untrans. fblntod calh and irr sorrre instances, protein can be completely elimina frttnt tho ntedlum (xee'l'uhle 4). 'l'ho ndvent of rnpid rncl sensitive nnulyticul mcthodrr (e,g, t{PLC) har
82

Whcn cells are grown as batch cultures, all the nutrients required for the drrrirtion of the culture must be present in the initial medium. The two major energy sources, glucose and glutamine, need to be present at unphysiologicnlly high concentrations which may lead to the production of high concentrallorrs of the toxic metabolites, lactate and ammonia. In the fed batch agrproach, glucose and glutamine can be added at intervals as they become tlcpleted, thus limiting toxicity and improving efficiency of utilization. ln perfusion systems, i.e. culture systems where cells are retained in llre f'ermenter while medium flows through, fresh medium is continuously
88

T. Cortwright ond G. P. Shoh

supplied and spent medium removed. Ideally, the perfusion rate and the centration of each component in the medium should be adjusted to match consumption rate of each nutrient. This requires detailed knowledge of cells' nutritional needs and of the rate of production of possible toxic products. Careful design of the medium is therefore needed to match mediu composition to the cells' metabolic needs. Another point that requires sideration is that perfusion rate should not be so rapid as to flush out au logous growth factors.

3: Culture medio
adcquate protection. [It will be found that highly purified trypsin is much less cflicient at releasing cells from surfaces than the impure preparations usually enrployed (45).1

5.2 Anchorage-dependent cells

Many of the cell types which are currently the focus of intense activity, such as endothelial cells and epithelial cells, will only grow function when attached to surfaces. Some of the cells used in vaccine ma facture, such as human diploid fibroblasts, Vero cells and MDCK cells, a exhibit anchorage dependence. The need for cells to be attached to surfaces imposes at least two requirements on the medium. Firstly, although some of these cells can
thesize their own attachment factors, attachment is generally accelerated viability improved if factors such as fibronectin and laminin, which form of the ECM, are incorporated in the medium formulation. These factors to the surface on which the cells will attach and act as ligands for specific

8.2.1 Aggregate cultures When cells that are normally anchorage-dependent are grown in conditions whcre attachment factors are limiting, they tend to form aggregates which can be propagated in suspension. Such cultures can be useful for vaccine production or other large-scale applications, but a limitation may be the heterogeneity of tbc cultures and a tendency for necrotic regions to develop in the centre of the larger aggregates with subsequent cell lysis and the liberation of cell contents end debris into the culture.

6.3 Stirred suspension cultures

Clnmage

A primary concern in suspension cultures is the protection required against to cells by shear forces. In serum-containing medium this protection

surfaca receptors called integrins. An alternative approach is to precoat the culture surface with ECM nents such as collagen or fibronectin (see Chapter 4, Section 5.4). Recently an engineered fibronectin substitute called Pronectin (available from Protei Polymer Technologies) has been developed for this purpose. This has t advantage of promoting better attachment than the natural matrix protei since it contains multiple copies of the sequence Arg-Gly-Asp recognized integrins, and also it avoids the need to add animal-derived material to culture system with the attendant risks of viral contamination (43).

lr supplied by serum proteins, particularly albumin. In serum-free conditions, rcvcral synthetic polymers (usually at around 1 g/litre) have been used to fill this role (see Section 4.2.5). llccent studies have shown that shear damage occurs when turbulent eddies ol'similar size to the suspended cells are produced in the liquid (46). This pltcnomenon appears to be more associated with bubble formation and collapsc such as is produced by cavitation or air entrainment caused by the lmpe ller, or by bubble disengagement from the surface when sparging, rather tltln by shear forces in the body of the liquid. Pluronic F-68 is particularly efl'cctive in protecting cells from this type of damage and is now widely used ln scrum-free medium as a protective agent against shear (see Section 4.2.5). l,imitation of oxygenation is the main factor which currently restricts the ;crlc and density of many animal cell cultures, and cell damage by sparging or

Polypeptide growth factors (particularly members of the FGF family hind to elements of the ECM which may act as a slow release pool thesc lactors. The dynamic interplay between growth factors, the matri runcl thc attached cells may be an important aspect of cellular physi
(44).

by high speed agitation precludes the use of these methods to increase ruxygenation within the culture, although improved physical protective agents
for the cells may improve this situation. An alternative approach which is widcly used is to perform oxygenation in a compartment of the bioreactor wltich is physically separated from the cells in order to avoid contact between cells and bubbles. This implies an efficient separation system and an adequate circulation system to ensure that oxygenated medium reaches all cells in the
lriorcactor. Irr some reactor configurations (some spin filter devices, hollow-fibre sysle rrrs) the cells are completely separated from the vessel in which oxygenation urcurs. In others such as encapsulated cells or cells in macroporous carriers, llrc oxygenation may occur in the same vessel as the cells, but they are ptrrtccted from direct contact with bubbles or liquid turbulence. An important consideration whcn Pluronic F-68 or other protective agents ere udded to products intendecl fbr therapeutic use is the need to remove

'l'lre sccond rcquirement that anchorage dependence imposes on medit ir the rrced to inhibit the proteolytic enzymes that are used for releasi eellr I'ront tlrc substrate during sub-culture in order to minimize cell damage, ln the pre$ence ol'$erum, this is achieved by the endogenous protease in. hitritoru. lrr cerunr-l'ree medium, protection can sometimes be achieved by rlnaing tho cellr with mcdium containing soya bean trypsin inhibitor after trypalrrllntion, lt nhoulel bo noted, however, thnt thc crucle 'trypsin' prep. uration corrrmorrly ured in ccll culture contuins proteolylic uctivities other than trypnln End thnl r singlc, purified protcasc lnhltritor nlry not provide
84

$[

T. Cortwright and G. P. Shoh

these effectively during downstream processing. Anti-foaming agents (particularly silicones) are sometimes added to cultures, but their use is best avoided when possible since such materials are notoriously difficult to eliminate completely from the final product.

3: Culture media
rul'adequate instrumentation to ensure that pH, pO2, and temperature remain

within the required limits.

5.4 High-density culture systems

As already indicated, several of the culture systems currently in use provide cells at veiy high density, often in excess of 108 cells/ml. This has advantages in that higher concentrations of product can be generated but imposes the need for better control of fermentation conditions to ensure that the cells are kept within the specified environmental limits. High cell density systems can be divided into two t5rpes, homogenous systems where the cells are continually mixed and plug flow systems where cells are held immobile while medium flows past them. Many configurations of these types of bioreactor exist-we will only consider two which represent the best performing examples of their
class.

6.4.2 Hollow-fibre systems ln hollow-fibre systems, the cells are held in the bioreactor separated from the mcdium flow by the walls of capillary tubes (hollow fibres) through which mcdium flows. Typically the capillary walls are impervious to macromolecules
but allow the passage of low molecular weight nutrients including oxygen (49).

In common with other plug flow systems, hollow-fibre bioreactors may

rul'l'er from the development of gradients of nutrients, particularly of oxygen, along the length of the medium flow path. Adequate medium flow rate is

to limit the heterogeneity of the culture that this may produce. Unlike the situation with macroporous carriers, profein products generated in hollow-fibre systems are not released into the bulk medium, but are retained by the capillary membrane in the cell compartment from where they can be periodically or continuously harvested.
fequired

5.4.1 Macroporous carrier systems

These systems employ porous particles which contain pores large enough to permit cells to enter the particle and to colonize the internal space. The particles can be of neutral buoyancy and used as microcarriers or they can be of higher density (typically 1.3-1.6 g/cm3) and employed in a fluidized bed configuration. It is this configuration which permits very efficient mass trans' fer, a homogenous cell culture, and high cell density (47). A particular feature of this and other high density systems is that the are able to form their own local micro-environment while still being able receive required nutrients and release toxic waste products to the bulk medium. An important consequence of this is the reduced need for grow factors and matrix components since these are secreted by the cells and maintained in their own micro-environment. In cases where cells do not naturally produce the required factors they may relatively easily be gineered to do so. ln addition, since cells inside the particles are not directly exposed to the bulk medium, protection against shear is much less of a problem. However, lhe protease levels secreted by cells in high density cultures may be significant

6.5 Very low density cultures

ekrning of cells requires that cellular proliferation is achieved at very low cell
derrsities.

Not surprisingly, here the opposite situation pertains to that dis-

eusscd above and cells require maximum support from the medium. Serum sttpplementation alone is often not sufficient and'conditioned medium' harvestcd from actively growing cells is frequently used. This contains undefined mncromolecules including growth factors and detoxifying factors, and low molccular weight compounds such as tricarboxylic acid cycle intermediates which may help with culture establishment. In extreme cases it may be necessary Itt use 'feeder layers' of metabolically active but non-proliferating cells (usually ptoduced by irradiation) which are co-cultured with the required cell (see eltrrpter 5, Section 2.4.4, and Chapters 6 and 7). More recently, efficient media for clonal growth have been developed wltich are based on standard media supplemented with recombinant growth

fsclors appropriate
se

to the required cells and with efficient pH

control

nrrd purticulur attention should be paid to the need to include protease lrrlribltorx to protcct product from degradation' ln hlgh.density cultures, medium may rapidly become depleted in particulur eotnlxtttentt und umino acid analysis of input and output medium streamg xlrould bc perlirrmecl to determine whether particular amino acids becomo llrnlting und lu perrnit ndjustment of medium concentration and/or flow uccordlngly, lt rhould lrc notcd that amino acid deficiency may be one factor, wlrieh crtn lncluco lhe prtxluction of proteolytic enzymer by nnimal cells (48), Another imporlnnt cunnitlcrution with high-dcnnity cultur$ in tho provision
80

lricved by the use of buffers such as Hepes or Mops. CO2 is also an essential rerlrrirement and since cells at low density may not produce enough to satisfy this, incubation in an atmosphere containing CO2at an appropriate concenIttrlion for the medium used is essential.

tl.

In-house medium development and production versus commercial supply 6.f Economic considerations
Mrtlium production from the basic raw materials is a very major undertaking Irrvolving as it does the sourcing und quality control of a large number of
$7

T. Cartwright ond G. P. Shoh

individual components. In addition, the technology involved in milling and mixing powders and ensuring homogeneity and compatibility of the different components in the mixture is complex and not likely to exist within most biotechnology companies or cell culture laboratories. For most users, the real choice to be made is between the different types of medium formulation and presentation that are available commercially. These include ready-to-use single strength (1x) medium, medium concentrates (10x or in some cases 50x) or pre-mixed powder. Liquid medium is also supplied in a variety of packaging ranging from small bottles (usually 5ffi ml) to flexible plastic containers with volumes from one to several hundred litres" Laboratory-scale operations are likely to favour bottled single strength medium. For small process operations, single strength medium in 10 or 20 litre bags is a particularly convenient approach. On larger scales the choice depends on technical and economic factors which are specific to each manufacturer. Use of large containers of single strength medium permits a minimum of investment in preparation, quarantine, and quality control facilities and equipment, and staff requirements may also be reduced. However, it is expensive both in terms of the litre cost of medium and possibly also in the cost of delivery. Storage of large medium volumes in acceptable conditions may also incur significant costs (note that this is also true for the storage during quarantine of medium prepared in-house). A compromise is the use of medium concentrates. In this case it is necessary to invest in plant for the production of tissue culture quality water and the vessels and pipework required for dilution. Continuous flow dilutors are available for on-line dilution. The cheapest approach in raw material costs is to buy powdered medium and prepare this for use on site. This requires considerable investment in an adequate medium kitchen with ancillary facilities and appropriate personnel to operate it. When evaluating these options, it is important from an economic viewpoint that the true cost of the installation and operation of the required medium kitchen facilities are fully evaluated. These include the installation itself, the
provision of adequate technical services, clean room facilities and disposables, truined operatives, and the necessary quality control backup.

3: Culture medio

6.3 Quality assurance and quality control

Clritical elements in quality assurance for medium are the choice of supplier, evaluation of the documentation provided by the supplier and audit of the ttupplier's facilities when appropriate. For basal and serum-free media, many

ruppliers have registered a drug master file with the relevant regulatory
nuthorities which provides the necessary guarantees concerning the manufacturing process. For serum, confidence must be based almost entirely on the bltch documentation provided. 'fo assure consistent performance of medium, standard procedures must be ret up in-house for all operations concerning the production of medium from hrught-in components. This includes standardization of the storage conditions and the duration of storage of the complete medium before use, since nany complete media have only limited stability. I-ocal quality control testing by the user before committing a medium batch trt a production process is still of critical importance. Testing of serum batches htts already been discussed in Section 3.5.1. Key tests to be applied to batches ol'liquid and powdered medium are summarized in Section 2.4.5. Simple tests such as measurement of osmolality and of pH should be applied to every Itrcdium preparation (including those from the same batch of powder or of concentrate) to ensure that no error in solution or dilution has been made. Sterility of the final medium before use is a less simple question because the

tirnc required for completion of the sterility test (7-10 days) may not be
eornpatible with storage of the complete medium which may be unstable over llris period. Some operators assume sterility if all preparation procedures hnve been performed without incident and then verify sterility in parallel with production. The speed and sensitivity of sterility testing can be improved by llrc use of filter concentration methods to detect very low levels of contaminrul i ng micro-organisms.

l{cferences l. llettger, W. and Ham,

0.2 Development of a dedicated medium

(hnrnrercinl dccisions regarding in-house development or sub-contracting to n crtrnrnerciul organization may also be important when optimization of urctliunr tirr t specilic cell type or process is required. Here the key question is how I'recptcrrtly it is necessary to develop a new dedicated medium and whather lho nruintcnance of in-house medium development expertise and fucllltier is curnonricully justified by this frequency. Several of the medium produetiun lruunes ol'l'er conlidential contract serviccs in which a team which is orrgaged lirll'tinrc in medium development will optimizc modium lbrmulatlonn r*peciflcally lirr lhe customer's cells.
88

R. (1982). Adv. Nutr. Res., 4,249. .1. ltoth,E,., Ollenschlager, G., Hamilton, A., Langer, K., Fekl, W., and Jaksez, R. ( 1988). In Vitro Cell Dev. Biol., ?A,96. .l . liagle, H. (1955). Science,122, 501. ,1 , Williams, G. T., Smith, C. A., Spooncer, E., Dexter, T. M., and Taylor, D. R. (1990). Nature, 34.3,76. I . Moses, H. L., Coffrey, R. J., Leof, E. B., Lyons, R. M., and Kesi-Oja, J. (1987).
.l

. Physiol., Suppl. 5, 1.

(r

Van der Pol, L. and Tramper,J. (1992).In Anirnalcelltechnology: developments, processes and produclr (ed. R. E. Spier, J. B. Griffiths, and C. MacDonald), pp.

7. llodgson, J. (1993). Biol'[echnoktgy, ll,49. It. f lrryashi, l. and Sato, A. (1976), Nature,2S9,
00

It)2-4. Buttcrworth-Hcinemann, Oxford.

132.