System. App!. Microbio!.

21, 539-545 (1998)

__G_us_ta_v_Fi_sc_he_rV_e_rla_g_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

SYSTEI\IL4TIC AND
APPLIED MICROBIOLOGY

Analysis of LMW RNA Profiles of Frankia Strains

by Staircase Electrophoresis
ENCARNA VELAzQUEZ!, EMILIO CERVANTES 2 , JOSE MARIANO IGUAL2, ALVARO PEIX 2 , PEDRO
SAAD BENAMAR3 , ANDRE MOIROUD\ CHRIS T. WHEELERS, JEFF DAWSON6 , DAVID LABEDA 7 ,
CLAUDINO RODRIGUEZ-BARRUEC0 2 , and EUSTOQUIO MARTINEZ-MoLINA!

F. MATEos\

Departamento de Microbiologia y Genetica, Edificio Departamental, Salamanca, Spain

IRNA-CSIC, Salamanca, Spain
3 Laboratoire de Biologie et Physiologie Vegetales et Forestieres, Ecole Normale Superieure, Bensouda, Pes. Maroc
4 Universite Lyon 1. Ecologie Microbienne, Villeurbanne Cedex, France
5 Division of Biochemistry and Molecular Biology, Institute of Biomedical and Life Sciences, Bower Building, University of Glasgow,
Scotland
6 Department of Natural Ressources and Environmental Sciences, University of Illinois, 1005a Plant Sciences Lab, Urbana, IL
7 National Center for Agricultural Utilization Research, 1815 North University, Peoria, IL USA
1

Received July 21,1998

Summary
An optimized technique of polyacrylamide gel electrophoresis, Staircase Electrophoresis (SCE), was applied to determine the stable Low Molecular Weight RNA (LMW RNA) profiles of 25 Frankia strains
from diverse geographic origins and host specificity groups as well as species from other actinomycete
genera. Application of the technique permits the rapid identification of Frankia strains and their differentiation from other actinomycetes. The isolates used in this study were grouped in eight clusters, each
comprising strains with identical LMW RNA profiles. Comparison of these results with others obtained
from DNA sequences or DNA hybridization methods suggest a high degree of complexity in the genus
Frankia. Application of SCE to profile LMW RNA should in the future facilitate biodiversity studies of
Frankia and discrimination of new species.
Key words: Frankia - LMW RNA - Nitrogen fixation - Bacterial taxonomy

Introduction
Bacteria belonging to the genus Frankia are slow
growing actinomycetes characterized by their capacity to
fix atmospheric nitrogen in symbiotic associations resulting in the formation of nodule structures in the roots of
many perennial woody dicotyledoneous plants.
Since the first reproducible report of isolation of a
Frankia strain in 1978 (CALLAHAN et a!., 1978) several
hundred of isolates have been obtained for study in diverse laboratories throughout the world and considerable progress has been made in understanding many aspects of the Biology of these polymorphic bacteria (BENSON and SYLVESTER, 1993). As for Rhizobium, the first attempts to establish a classification of the genus Frankia
were based on their host range (BEeKING, 1970) and diverse phenotypic characteristics (LALONDE et a!., 1988).
More recently, various techniques, based on molecular
biology protocols, have been used for this purpose.

Studies of DNA relatedness, based on DNA-DNA reassociation kinetics, revealed the existence of at least
nine genomic species (FERNANDEZ et a!., 1989). Among
these, three contained strains compatible with the Alnus
specificity group, five with the Elaeagnaceae and one
with Casuarina. Genomic species 1 was proposed to be
Frankia alnii, the type species of the genus. These studies
are hampered by the difficulty to isolate and grow some
strains in laboratory media as well as by the low efficiency of DNA extraction from Frankia cultures.
Other approaches to allocating Frankia strains to taxonomic groups were based on the comparative analysis
of PCR amplified sequences, an approach which also fa-

Abbreviations: LMW RNA - Low Molecular Weight RNA;

SCE - Staircase Electrophoresis

540

E. VELAzQUEZ et al.

cilitated the characterization of uncultured Frankia

strains, for which template DNA was directly obtained
from the nodules. NAZARET et al. (1991) developed a
protocol based on DNA amplification and sequencing of
the 268 base pairs corresponding to partial ribosomal
DNA sequences among eight of the previously described
genomic species. Their work included the elaboration of
a phylogenetic tree, showing a close relatedness among
strains belonging to the Casuarina and Alnus infectivity
groups, with both groups well separated from the
Elaeagnus
infectivity
group.
Atypical
strains.
ORS020602 (Dll) and 55005 (DBB02060510) isolated
from, but unable to reinfect Casuarina plants, were
grouped close to Elaeagnus infective strains. In general,
strains from Elaeagnus groups showed less diversity than
those in the Alnus groups.
From an analysis of PCR amplified sequences corresponding to 23S rRNA, HONNERLAGE et al. (1994) described seven groups, four related to the Alnus compatibility group, one including strains related to Casuarina
compatibility group, another including Elaeagnus related
strains and finally, a group consisting exclusively of the
uncultured endophyte of Coriaria nepalensis. These results do not match exactly the results of the DNA-DNA
reassociation experiments. Thus, different strains separated by a short distance in the phylogenetic tree of HONNERLAGE et al. (1994) are sometimes included in different
genomic species by FERNANDEZ et al. (1989), whereas
strains separated by longer distances are sometimes
placed together. However, a similar grouping of strains as
related to their host compatibilities is observed.
Based on the complete nucleotide sequences of the 16S
rRNA of eight Frankia strains, others in databases and
DNA sequences amplified from nodules, NORMAND et al.
(1996) described the clustering of Frankia into four
groups: Cluster 1 included strains that infect Alnus and
Casuarina species, Cluster 2 included the unisolated symbiont of Dryas, related to the unisolated strains from Coriaria and Datisca, Cluster 3 included Elaeagnus infective Frankia strains and Cluster 4 included unclassified
strains Pdl and atypical strains Cn7, Dc2 and AgBl-9.
Given the present status of Frankia taxonomy, alternative methods may make an important contribution to
the classification of the genus and determination of the
phylogenetic relationships among diverse strains. BEYAZOVA and LECHEVALIER (1992) used LFRA to analyze the
phylogenetic relationship among more than 100 Frankia
strains. This technique is not particularly easy to master
in order to obtain satisfactory and reproducible analysis
and efforts to optimize other, simpler and faster methods,
may results in significant advances in these studies.
HOFLE (1988) proposed to use the profiles of Low
Molecular Weight, stable RNA (LMW RNA) in studies
of bacterial taxonomy. The LMW RNAs include the 5S
rRNA and class 1 and 2 of tRNA (HOFLE, 1988). Recently, a new electrophoretic technique, Staircase Electrophoresis (SCE), has been developed that allows optimal separation of these molecules and their utilisation in
bacterial taxonomy (CRuz-SANCHEZ et aI., 1997). This
new technique has been applied to the differentiation of

species in the family Rhizobiaceae, where different LMW

RNA profiles have been found to correspond well with
the described genera and species (VELAZQUEZ et aI.,
1998). In the present work, the LMW RNA profiles of
Frankia isolates from diverse host plants and from diverse geographic provenances have been analysed.

Materials and Methods

Bacterial strains and media: A description of the Frankia
strains used in this study including the host plant from which
they were derived and their geographic origin is shown in Table
1. Isolation of new strains for the present work was carried out
by published protocols (LECHEVALIER and LECHEVALIER, 1990).
Other actinomycetes obtained from culture colections are
Rhodococcus rhodochrous CECT3046 (ATCC4273), Dactylosporangium aurantiacum CEcn288 (ATCC2391), Streptomyces cinammoneus (former Streptoverticillium cinammoneus)
CECT3258 (ATCCI11874), Streptomyces kentuckense (former
Streptoverticillium kentuckense) CEcn262 (ATCC12691),
Micromonospora melanosporea CECn087 (CBS 270.62),
Streptomyces halstedii NRRL-2381 and Streptomyces lividans
JI1326. The Frankia strains used in this study were cultivated in
Qmod medium (LALONDE and CALVERT, 1979) without glucose
and with pyruvate at 0,5%. The cultures were kept at 25C for
three weeks. The other actinomycete strains were grown in
YEG medium (Yeast Extract 0,7%, Glucose 1 %) at 25C for a
week.
RNA extraction and LMW RNA profile analysis: The RNA
of the strains studied was extracted as described by HOFLE
(1988). LMW RNA profiles were obtained using Staircase Electrophoresis in 14% polyacrylamide gels under denaturing conditions in steps of 10 min, rising through a constant ramp with
50 V increases from 100 V to 2300 V as reported earlier (CRUZSANCHEZ et aI., 1997). The following commercial molecules
from Boehringer Manheim (Manheim, Germany) and Sigma
(St. Louis, MO, USA) were used as reference: 5S rRNA from
Escherichia coli MRE 600 (120 and 115 nucleotides) (BIDLE
and FLETCHER, 1995), tRNA specific for tyrosine from E. coli
(85 nucleotides) and tRNA specific for valine from E. coli (77
nucleotides) (SPRINZL et aI., 1985). Samples were prepared as
reported elsewere (CRuz-SANCHEZ et aI., 1997). After electrophoresis, the gels were silver-stained as described by HAAS et
al. (1994).
Data analysis and construction of dendrograms: The bands
present in each profile were coded for input into a data base
that included all the strains studied and Jaccard's similarity
coefficient was calculated to construct the distance matrix. A
dendrogram was constructed from the distance matrix using the
Unweighted pair Group Arithmetic Mean (UPGMA).

Results
RNA extraction and LMW RNA profile analysis
The LMW RNA profiles of the strains used in this
study are shown in Figure 1. All contain the three expected zones: 5S RNA, class 2 tRNA and class 1 tRNA
(CRuz-SANCHEZ et aI., 1997). The number of bands present in the tRNA zone was as expected according to previous results obtained for the E. coli strain CECT99
(ATCC9637), using Staircase Electrophoresis (CRUZ-

LMW RNA from Frankia

nt

\1\-\

nt

120

20

115

115

85

85

77

77

)IW

541

Fig. 1. LMW RNA profiles of the strains of actnomycetes used in the present work.
(A) All lanes correspond to Frankia. Lane 1: Cn3 from Coriaria nepalensis. Lane 2: Cn7 from Coriaria nepalensis. Lane 3: Hr 114.2
from Hippophae rhamnoides. Lane 4: Hr 77.3 from Hippophae rhamnoides. Lane 5: Ag 67.5, Ag 95.2, Ar 112.2, AcN14A,
ENS010712, ENS010714, UGLOl1301, UGL013103, UGL013104, UGL010701, NRRLB-16510, all strains from different species
of Alnus, NRRLB-16510 (HrI1) from Hippophae rhamnoides. Lane 6: CS1020602, CS1020604, CS1020620, CSSI020621,
UGL020603, all strains from Casuarina equisetifolia. Lane 7: Strains 55005, ORS020602, atypical from Casuarina and Strain Ea
32.1 isolated from Elaeagnus. Lane 8: Strains NRRLB-16422 (MgI8) and NRRLB-16423 (Pd1) isolated respectively from Myrica
gale and Purshia tridentata.
(B) All lanes correspond to other actinomycetes: Lane 1: Rhodococcus rhodochrous CECT3046 (ATCC4273). Lane 2: Dactylosporangium aurantiacum CECT3288 (ATCC23491). Lane 3: Micromonospora melanosporea CECT3087 (CBS270.62). Lane 4: Streptomyces cinammoneus CECT3258 (ATCC11874). Lane 5: Streptoverticillium kentuckense CECT3262 (ATCC12691). Lane 6:
Streptomyces halstedii NRRLB-2381. Lane 7: Streptomyces lividans JI1326.

SANCHEZ et a!., 1997) and for the members of the Rhizobiaceae (VELAzQUEZ et a!., 1998).
The Frankia strains used in this study are distributed
in eight groups, each showing a different LMW RNA
profile (Figure 1A). Groups 1 and 2 include strains isolated from Coriaria nepalensis: strain Cn3 (Fig. lA, lane
1) and Cn7 (Fig. lA, lane 2). Groups 3 and 4 include
strains isolated from Hippophae rhamnoides: Hr1l4.2
(Fig. lA, lane 3) and Hr77.3 (Fig. lA, lane 4), Group 5
(Fig. lA, lane 5) includes strains UGL010701,
UGL013l04, UGL013l03, UGLOl1301, ENS010712,
ENS010714, AcN14A, Ag67.5, Ag95.2 and Ar112.2, all
of them isolated from diverse species of Alnus. Group 6
(Fig. lA, lane 6) includes atypical Frankia strains isolat-

ed from Casuarina that infect Elaeagnus (ORS020602

and 55005) as well as strains Ea32.1 isolated from
Elaeagnus angustifolia. Group 7 (Fig. lA, lane 7) includes strains CSI020602, CSI020604, CSI020620,
CSI020621 and UGL020603 isolated from Casuarina
equisetifolia. Group 8 (Fig. lA, lane 8) includes strains
NRRL-B16422 (MgI8) isolated from Myrica gale and
NRRLB-16423 (PtI1) isolated from Purshia tridentata.
The LMW RNA profiles of other actinomycetes are
shown in figure 1 B: Rhodococcus rhodochrous
CECT3046 (Lane 1). Dactylosporangium aurantiacum
CECT3288 (Lane 2), Micromonospora monosporea
CECT3087 (Lane 3), Streptoverticillium cinammoneum
CECT3258 (Lane 4), Streptoverticillium kentuckense

542

E.

VELAzQUEZ

et al.

CECT3262 (Lane 5), Streptomyces halstedii NRRLB2382 (Lane 6) and Streptomyces lividans JI1326 (Lane 7).

The Frankia strains used in this study are separated in

two groups with a similarity coefficient of 0.4. One
group includes strains NRRLB-16422 (isolated from
Myrica gale) and NRRLB-16423 (isolated from Purshia
tridentata), both strains with the same LMW RNA profile (Fig. lA, lane 8). The second group is subdivided in
two with a coefficient of 0.5. One of these subgroups includes two strains isolated from Coriaria, CN3 and CN7
(Figure lA, lanes 1 and 2). The remaining Frankia strains
are divided in two groups with a similarity coefficient of
0.6. One of them comprises the two strains obtained
from Hippophae rhamnoides (Hr 114.2 and Hr 77.3),
the other is further subdivided into two groups with a
similarity coefficient of 0.8. One of these comprises sev-

Data analysis and construction of dendrograms

Construction of a dendrogram (Figure 2) revealed two
major groups with a similarity coefficient of only 0.24.
One contains Micromonospora and Dactylosporangium
and the other group contains the rest of the actinomycetes (including Frankia strains) studied in this work.
This second group is further divided in two clusters with
a similarity coefficient of 0.3, one contammg all the
Frankia strains and the other the remaining actinomycetes.

CN3

CN7

r--

Hr1l4.2
Hr77.3

UGLOI0701
UGL013103
UGLOI3104
UGLOIl301
ENSOI07l2
ENSOI0714
SI4
AcNI4A

Ag67.5
Ag9S.2

Arll2.2

ORS020602
55005
Ea32.1
CS1020602
CS1020604
CS1020620
C81020621
UGL020603
816422

r--

RhodocOCCIIS rhodocrholls

I
I

0.2

816423

0.5

Streptoverticillium cinamlllOnel/1Il
StreptoverlicilliulII Kenll/keme
Streptomyces haldstedii
Streptolllyces IMdans
Dactilosporangium aurantiacum
Micromonospora lIIonosporea

Fig. 2. UPGMA dendrogram

based on Jaccard's coefficient
derived from LMW RNA
profile characteristics for the
actinomycete strains.

LMW RNA from Frankia

eral strains isolated from Casuarina equisetifolia

(UGL020603) and the isolates obtained in Salamanca
CS1020602, CS1020620 and CSI020621) . The remaining
strains are divided between a group that includes all the
tested strains from Alnus and a final group with two
atypical strains from Casuarina and strain Ea32.1 from
Elaeagnus (similarity coefficient 0.86).

Discussion
In recent years new techniques based on molecular biology protocols have been applied to the resolution of
traditional questions in microbial taxonomy. The well
known difficulties for isolating and growing in culture
Frankia strains have led to the application of techniques
which do not require previous steps of isolation and culture of microbial strains in the laboratory (HONNERLAGE
et al., 1994; NAZARET et al., 1991; NORMAND et al.,
1996). Taxonomic studies of this genus have shown in
general the relatedness among strains isolated from the
same host plant, but there are controversial results concerning the relative taxonomic position in dendrograms
of strains isolated from diverse host plants.
As for other symbiotic, nitrogen-fixing bacteria,
namely those in the family Rhizobiaceae, attempts were
made initially to make the definition of species on the
bases of the capacity to nodulate a particular host plant
(BECKING, 1970). This character in itself is not a criterium robust enough for classification (TORREY, 1990) because particular plants may be nodulated by strains that
were derived from different host species or even genera.

543

In this sense, promiscuous hosts are Elaeagnus angustifolia and Alnus glutinosa. Also, most Frankia strains are
able to reinfect individuals of the host plant species from
which they were isolated. Nevertheless there are reports
of so called atypical strains, i. e. strains unable to infect
the host plant from which they are derived but able to infect other actinorhizal genera (BAKER, 1987). These results emphasize the need for the application of new taxonomic techniques, that in the Rhizobiaceae were decisive
for the description of new genera and species.
The application of the technique of Staircase electrophoresis to obtain the LMW-RNA profiles in the Rhizobiaceae resulted in the reported observations that different genera show differences in their 5S RNA whereas
different species of the same genus show differences in
the tRNAs and overall, the LMW-RNA profiles are characteristic of each species. Moreover, the analysis of the
corresponding data by UPGMA resulted in similar dendrograms as those obtained by 16S rRNA sequencing
and analysis in these bacteria (VELAZQUEZ et al., 1998).
Based on these considerations, the LMW RNA profiles of Frankia strains from diverse host plants and geographic provenances were analysed by Staircase Electrophoresis. The technique was simultaneously applied
to other actinomycetes, showing different LMW RNA
profiles among the diverse species and when compared to
Frankia isolates (Figure 1). The clustering of actinomyceta I species was similar to the obtained with 16S rRNA
sequence analysis (GOODFELLOW, 1989), thus Dactylosporangium was more related to Micromonospora,
being both separated from the other groups. Rhodococcus remained isolated from the species of the genus

Table 1. Characteristics of Frankia strains used in this study.

Designation

Host source

Location

Source

Reference

NRRLB-16422 (MGI8)
NRRLB-16423 (PtIl)
ENS010712 (S12)
ENS010714 (S14)
UGLOI0701
UGL013103
UGL013104
UGL011301
UGL020603
Ea32.1
Hr114.2
Hr77.3
Ag 67.5
Ag 97.2
Ar 112.2
ACN 14a
ORS020602 (Dl1)
Cn3
Cn7
55005 (DBB02060510)
CSI020602
CSI020604
CSI020620
CSI020621

Myrica gale
Purshia tridentata
Alnus glutinosa
Alnus glutinosa
Alnus glutinosa
Alnus rubra
Alnus rubra
Alnus inokumai
Casuarina equisetifolia
Eleagnus angustifolia
Hyppophae rhamnoides
Hyppophae rhamnoides
Alnus glutinosa
Alnus glutinosa
Alnus rubra
Alnus cordata
Casuarina equisetifolia
Coriaria nepalensis
Coriaria nepalensis
Casuarina equisetifolia
Casuarina equisetifolia
Casuarina equisetifolia
Casuarina equisetifolia
Casuarina equisetifolia

USA
USA
Morocco
Morocco
Scotland
Scotland
Scotland
S. Korea
Egypt
Francia
Francia
Francia
Francia
Francia
Francia
Canada
Senegal
Pakistan
Pakistan
USA
Spain
Spain
Spain
Spain

D. Labeda
D . Labeda
S. Benamar
S. Benamar
C. T. Wheeler
C. T. Wheeler
C. T. Wheeler
C. T. Wheeler
C. T. Wheeler
A. Moirud
A. Moirud
A. Moirud
A. Moirud
A. Moirud
A. Moirud
P. Normand
H. Ramirez
H. Ramirez
H. Ramirez
M.Igual
this study
this study
this study
this study

BAKER, 1987
BAKER, 1987
(Unpublished data)
(Unpublished data)
WHEELER et al., 1986
HOOKER and WHEELER, 1987
HOOKER and WHEELER, 1987
SAYED et al., 1997
(Unpublished data)
(Unpublisehd data)
(Unpublished data)
(Unpublished data)
(Unpublished data)
(Unpublished data)
(Unpublisehd data)
NORMAND and LALONDE, 1982
GAUTHIER et al., 1981
MIRZA et al., 1992
MIRZA et al., 1992
BAKER, 1987

544

E. VELAZQUEZ et al.

Streptomyces that clustered toghether confirming the

finding of WITT and STACKEBRANDT (1990). Our results
support the unification of the genera Streptoverticillium
and Streptomyces as proposed by these authors because
their 5S rRNA show an identical profile.
Among the Frankia strains, eight LMW RNA profiles
were observed that, in general, correspond to groups of
strains isolated from different host plants, although
strains with different LMW RNA profiles may be obtained from the same host plant (Figure lA, lanes 1 to 4)
and strains with the same LMW RNA may be obtained
from different hosts (Figure lA, lane 8).
Strains isolated from Alnus consistently showed the
same LMW RNA profile independently of the host
species and their geographic origin (Table 1). Strain
ACN14a was in cluster 1 of NORMAND et a!. (1996) and
closely associated to CpIl, the strain designated as the
type strain for Frankia alnii, in the analysis of HONNERLAGE et a!. (1994). The uniformity in the LMW RNA
profiles obtained for the Alnus strains, including
ACN14a, supports the designation of Frankia alnii as the
species type of the genus and this technique provides a
rapid and efficient method to identify strains belonging
to this species.
The typical strains isolated from Casuarina that were
used in this study have an identical LMW RNA profile
and constitute the same species. Atypical strains, isolated
but unable to re-infect Casuarina, but able to infect
Elaeagnus (BAKER, 1987; NAzARET et a!., 1991) 55005
and ORS020602 share their LMW RNA profile with
strain Ea32.1 isolated from Elaeagnus and may thus be a
separate species.
The two Frankia species comprising the Alnus isolates
and the typical Casuarina isolates are close to each other
phylogenetic ally, in agreement with the results of other
authors (HONNERLAGE et a!., 1990; NAZARET et a!., 1991;
NORMAND et a!., 1996).
The low similarity coefficient among many clusters of
Frankia strains makes it likely to consider the existence
of different bacterial species (or even genera) in the genus
Frankia. This possibility is supported by the observed
variability in the region corresponding to 5S rRNA. Several authors have shown before that differences in the 5S
rRNA reflect differences among bacterial genera (HOFLE,
1988; HOFLE, 1990; VELAzQUEZ et a!., 1998).
Strains Hr114.2 and Hr77.3 isolated from Hippophae
grouped in the same cluster but the coefficient of similarity (0,8) indicates that each of the strains may belong to a
different species.
Strains Cn3 and Cn7, isolated from Coriaria, grouped
separately from the other Frankia strains. Similar results
were obtained for strain Cn7 by RAMIREZ-SAAD et a!.
(1998) using sequence data comparisons. Our results,
showing a low coefficient of similarity between Cn3 and
Cn7 (0.6), probably indicate that both strains may belong to different species.
In summary, our results show that the grouping of
Frankia isolates in taxonomic groups broadly reflects the
host plants from which they were derived. This conclusion confirms previous analyses based on diverse tech-

niques (BEYAZOVA and LECHEVALIER, 1992; FERNANDEZ et

aI., 1989; LALONDE et aI., 1988; NAZARET et aI., 1991;
NORMAND et a!., 1996) but presents interesting exceptions that may be grouped in two classes: 1) Strains from
very divergent host plants that share a very similar or
identical LMW RNA profile, as for example MgI8 and
Pd1 and 2) Strains isolated from the same host plant that
have very divergent LMW RNA profiles, as several isolates obtained from Coriaria and Hippophae in our
study.
The results presented indicate the usefulness of LMW
RNA profiles to differentiate species in the genus Frankia
and to establish taxonomic groups among strains isolated from diverse host plants and from diverse geographic
origins. The technique will allow in the future easy and
rapid identification of Frankia strains and their allocation to the described species of the genus. This will improve significantly knowledge of the taxonomy of this
bacterial group and help in studies of the biodiversity of
these ecologically and economically important bacteria.
Acknowledgements
We thank PHILIPPE NORMAND, HUGO RAMIREZ SAAD and
EMETERIO IGLESIAS JIMENEZ for suppling Frankia strains and
their commentaries to this work. This work was supported by
Grant CS101-97 of Junta de Castilla y Leon and the DGICYT
(Direccion General de Investigacion Cientffica Tecnica) Grant
PB92/097.

References
BAKER, D. D.: Relationships among pure-cultured strains of
Frankia based on host specificity. Physiol. Plant. 70,
245-248 (1987).
BECKING, J. H.: Frankiaceae fam. nov. (Actinomycelates) with
one new combination and six new species of the genus
Frankia Brunchorst 1886. Int. J. Syst. Bact. 20, 201-220
(1970).
BENSON, D. R. and SILVESTER, W. B.: Biology of Frankia strains,
Actinomycete symbionts of actinorhizal Plants. Microbiol.
Rev. 57,293-319 (1993).
BEYAZOVA, M. and LECHEVALIER, M. P.: Low-frequency restriction fragment analysis of Frankia strains (Actinomycetales).
Int. J. Syst. Bacteriol. 42, 422-433 (1992).
BIDLE, K. D. and FLETCHER, M.: Comparison of free-living and
Particle-associated bacterial communities in the Chesapeake
bay by stable Low Molecular Weight RNA analysis. Appl.
Environm. Microbiol. 61, 944-952 (1995).
CALLAHAN, D., DEL TRIDICI, P., TORREY, J. G.: Isolation and cultivation in vitro of the actinomyete causing root nodulation
in Comptonia. Science 199, 899-902 (1978).
CRuz-SANCHEZ, J. M., VELAZQUEZ, E. MATEOS, P. E, MARTINEZMOLINA, E.: Enhancement of resolution of Low Molecular
Weight RNA profiles by Staircase Electrophoresis. Electrophoresis 18, 1909-1911 (1977).
FERNANDEZ, M. P., MWGNIER, H., GRIMONT, P. A. D., BARDIN,
R.: Deoxyribonucleic acid relatedness among members of the
genus Frankia. Int.]. Syst. Bacteriol. 39,424-429 (1989).
GAUTHIER, D., DIEM, H. G., DOMMERGUES, Y. R.: In vitro nitrogen fixation by two actinomycetous strains isolated from
Casuarina nodules. Appl. Environ. Microbiol. 41, 306-308
(1981).

LMW RNA from Frankia

GOODFELLOW, M.: The Actinomycetes I, pp. 2333-2339. In:
Bergey's Manual of systematic bacteriology. vol. 4 (N. R.
KRIEG and J. G. HOLT eds.). 1st ed., Baltimore, Williams and
Wilkins 1989.
HAAS, H. B., BUDOWLE, B., WEILER, G.: Horizontal polyacrylamide gel electrophoresis for the separation of DNA framents. Electrophoresis 15, 153-158 (1994).
HOFLE, M. G.: Identification of bacteria by low molecular
weight RNA profiles: a new chemotaxonomic approach. J.
Microbiol. Meth. 8,235-248 (199).
HOFLE, M. G.: Transfer RNA as a genotypic fingerprinting of
eubacteria. Arch. Microbiol. 153,299-304 (1990).
HONNERLAGE, W., HAHN, D., ZEPP, K., ZEYEER, ]., NORMAND,
P.: A hypervariable 23S rRNA region provides a discriminating target for specific characterization of uncultured and cultured Frankia. Syst. Appl. Microbiol. 17,433-443 (1994).
HOOKER, J. E., WHEELER, C. T.: The effectivity of Frankia for
nodulation and nitrogen fixation in Alnus rubra and A. glutinosa. Physiol. Plant. 70, 333-341 (1987).
LALONDE, M., CALVERT, H. E.: Production of Frankia Hyphae
and spores as an infective inoculant for Alnus species, pp.
95-110. In Symbiotic Nitrogen Fixation in the management
of temperate forests (J. C. GORDON, C. T. WHEELER and D.
A. PERRY, eds.). Corvallis, Oregon, for. Res. Lab. Oregon
State University 1979.
LALONDE, M., SIMON, L., BOUSQUET, ]., SEGUIN, A.: Advances in
the taxonomy of Frankia: Recognition of species alnii and
elaeagnii and novel subspecies pomerii and vandijkii, pp.
671-680 In Nitrogen fixation: Hundred years after (H.
BOTHE, E DEBRUI]N and W. E. NEWTON, eds.), Stuttgart, Germany, Gustav Fischer Verlag 1988.
LECHEVALIER, M. P. and LECHEVALIER, H. A.: Systematics, Isolation and Culture of Frankia, pp. 35-60 In The Biology of
Frankia and Actinorhizal Plants (c. R. SO-IWINTZER and J. D.
T]EPKEMA, eds). Academic Press, 1990.
LUMINI, E., Bosco, M., FERNANDEZ, M. P.: PCR-RFLP and total
DNA homology revealed three related genomic species
among broad host range Frankia strains. FEMS Microbiol.
Ecol. 21, 303-311 (1996).
MIRZA, M. S., HAHN, D., AKKERMANS, A. D. L.: Isolation and
characterization of Frankia strains from Coriaria nepalensis.
Syst. Appl. Microbiol. 15,289-295 (1992).
NAZARET, S., COURNOYER, B., NORMAND, P., SIMONET, P.: Phylogenetic relationships among Frankia genomic species deter-

545

mined by use of amplified 16S rDNA sequences. J. Bacteriol.

173,4072-4078 (1991).
NORMAND, P., LALONDE, M.: Evaluation of Frankia strains isolated from provenances of two Alnus species. Can. J. MicrobioI. 28, 1133-1142 (1982).
NORMAND, P., ORSO, S., COURNOYER, B., JEANNIN, P.,
CHAPELON, c., DAWSON, J., EVTUSHENKO, L., MISRA, A. K.:
Molecular phylogeny of the genus Frankia and related genera and emendation of the family Frankiaceae. Int. J. Syst.
Bacteriol. 46,1-9 (1996).
RAMIREZ-SAAD, H., JANSE, J. D., AKKERMANS, A. D. L.: Root
nodules of Ceanothus caeruleus contain both N 2-fixing
Frankia endophyte and a phylogenetically related Nod-lFixactinomycete. Can. J. Microbiol. 44,140-148 (1998).
SAYED, W. E, WHEELER, C. T., ZAHRAN, H. H., SHOREIT, A. A.
M.: The effect of temperature and soil moisture qn the survival and symbiotic effectivity of Frankia. J. Soil BioI. Fert.,
in press (1997).
SPRINZL, M., MOLL, J., MEISSNER, E, HATMANN, T.: Compilation of tRNA sequences. Nucleic Acid Res. 13, 1-49 (1985).
TORREY, J. G.: Cross-Inoculation Groups within Frankia, pp
83-106 in The Biology of Frankia and Actinorhizal Plants
(c. R. SCHWINTZER and J. D. T]EPKEMA eds.). Academic
Press, 1990.
VELAZQUEZ, E., CRuz-SANCHEZ, J. M., MATEOS, P. E and
MARTINEZ-MoLINA, E.: Analysis of stable Low-Molecular
Weight RNA Profiles of Members of the Family Rhizobiaceae. Appl. Environ. Microbiol. 64, 1555-1559 (1998).
WHEELER, C. T., HOOKER, J. E., CROWE, A., BERRIE, A. M. M.:
The improvement and utilization in forestry of nitrogen fixation by actinorhizal plants with special reference to Alnus in
Scotland. Plant Soil 90, 393-406 (1986).
WITT, D., STACKEBRANDT, E.: Unification of the genera Streptoverticillum and Streptomyces, and amendation of Streptomyces Waksman and Henrici 1943, SO: Syst. Appl. Microbi01. 13,no.4,361-371 (1990).

Corresponding author:
EMILIO CERVANTES, IRNA-CSIC, Apartado 257, 37080 Salamanca, Spain
Phone: 011-34-923-219606, Fax: 011-34-923-219609