Beruflich Dokumente
Kultur Dokumente
FEBS 29829
Abstract The relation between the position of mutations in Saccharomyces cerevisiae metabolic network and their lethality is
the subject of this work. We represent the topology of the network by a directed graph: nodes are metabolites and arcs represent the reactions; a mutation corresponds to the removal of all
the arcs referring to the deleted enzyme. Using publicly available
knock-out data, we show that lethality corresponds to the lack of
alternative paths in the perturbed network linking the nodes
aected by the enzyme deletion. Such feature is at the basis of
the recently recognized importance of marginal arcs of metabolic networks.
2005 Federation of European Biochemical Societies. Published
by Elsevier B.V. All rights reserved.
1. Introduction
Information regarding genomes and metabolic interactions
are stored in many databases and scientists have the urgent
need for new tools in order to nd eective representations
and models for such interaction networks. Only a multidisciplinary approach may lead to signicant results, which implies
the merging of biological knowledge and methodologies coming from many and diverse elds such as mathematics, statistics, control theory, computer science, etc.
In the present work, we adopted a simple heuristic principle
in use for the study of failures in power supply networks: the
eects of a damage (in the form of the elimination of a preexistent link) on the entire network are related to the fact that
elimination of the arc isolates one or more node from the rest
of the network. That is to say that failures with greater consequences on the entire system are those for which it is not possible to devise alternative pathways for the nodes aected by
the arc(s) elimination. This simple principle is a very old topic
of biochemists when considering the eect of a step elimination
in biochemical networks represented as directed graph having
metabolites as nodes and reactions as arcs. The principle
implies that essential mutations, corresponding to the enzymes
whose elimination ends up in the death of the organism, are
*
Corresponding author. Fax: +39 064 9902355.
E-mail address: alessandro.giuliani@iss.it (A. Giuliani).
0014-5793/$30.00 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.febslet.2005.07.033
3. Results
A metabolic network can be imagined as composed by four
distinct classes of nodes (Fig. 2). The rst one is the strong
connected components (SCC), made of nodes each other linked
by directed paths; the reactant subset consists of metabolites
entering the system as reactants and called sources because they
can reach the SCC, but cannot be reached from it. The third class
is represented by the product subset, made of metabolites that
are accessible from the SCC, but dont have any connections to
it. Such metabolites are positioned at the end of a given pathway
and thus exiting the system as products (sinks). The metabolites
in the last class, called isolated subset, are inserted in autonomous pathways with respect to SCCs.
This connectivity structure can be found even in the architecture of scale-free networks such as Word Wide Web, the electric power transmission grid or airport connections [610].
4643
Fig. 2. General representation of a metabolic network, adapted from [18]. On the top, the SCC module corresponds to strongly connected
components; on the left there is the reactant subset (sources); on the right, the product subset (sinks); the isolated subset is on the bottom.
4644
463
332
356
418
1107
2350
448
2504
26
5715
129
685
3287
341
1251
3912
1250
232
334
169
327
2637
84
1169
337
475
112
295
3508
41
106
5824
5684
5512
360
5698
5692
262
131
206
73
409
1137
4051
4734
19
21
655
130
212
104
224
1260
20
575
3794
385
81
35
144
242
1762
330
2739
3090
4640 4376 3838
3373
294
147
8
620
942
361
4916
2741
440
9332
921
1300
143
101
4874
4895
286
718
4896
568
1304
44
29
1236
119
4751
103
92
345
199
117
4677
5401
446
257
301
1031
5402
5397
52
31
1177
121
1419
3899
51
4823
155
4425
559
281
669
1118
5699
447
124
689
231
6007
3451
1077
1528
1613
492
89
310
6008
6005
97
5688
364
279
6006
188
1102
459
1235
5404
5399
1083
186
979
2291
214
95
3465
22
109
881
363
137
5400
794
256
65
37
2972
5261168
239
118
5714
1242
380
365
1194
407
263
6459
1245
159 4549
275
85
1005
1103
2051
1277
422
2218
931
441
1346
460
5520
430
105
15
96
1885
636
197
75
63
1159
3232
299
705
4637
354
4691
631
579
6157
527
3082
55
5674
269
236
438
458
5676
2737
43
665
74
1136
322
357
352
944
152
49
4256
329
469
36
64 555
641
6156
248
3406
62
111
2652
5675
1241
350
230
3939
5673
570
3892
184
493
5665
4618
3557
4233
346
93
222
3175
91
42
437
750
1269
99
517
1233
6893
33
5744
2232
5829 4002
4133
251
900
624
77
328
4039
6244
158
5662
25
1165
134
4181
189
1094
4230
157
1120
116
183
588
670
577
254
122
148
235
315
1209
24
114
416
1179
108
141
417
166
966
83
311
1143
681
258
4302
2631
1271
149
48
5572
307
82
522
123
603
79
1302
4411
5354
3428
826
3506
864
499
1054
751
233
1551
1724
4236
3479
415
4807
445
234
504
5928
362
Fig. 3. Nodes coloured in blue, light blue, violet and green belong to the SCCs while red nodes pertain to reactant, product and isolated subsets.
Red arrows show the location of essential enzymes. For a better view, see the eps gure in the website http://www.dis.uniroma1.it/~farina/Yeast/, the
excel les reporting the information regarding the gure are enclosed.
4. Discussion
The results of our work are consistent with the recently discovered relevance of the so called non-hub-connectors [11].
These enzymes are located at the periphery of the metabolic
modules and connect dierent functionalities into a coordinated whole. This is in somehow reminding the structure of
power supply networks, where the most crucial electrical lines
are those transferring electricity from one strongly connected
cluster to another. It is important to stress the fact that our
representation of the metabolic network is the classical one
having the metabolites as nodes and the enzymes as arcs. Thus
the modications of the network correspond to the elimination
of a connection and not of a specic element (node). At odds
with this formalization, the systems biology approaches based
on protein networks are based on the elimination of specic
nodes: in these representations [1214] the proteins are the
nodes and the edges correspond to the presence of an interaction between the two intervening proteins.
In the proteinprotein interaction case, the central position
of a network element is not a signature of failure resistance,
4645
Fig. 4. The gure reports the fatty acid biosynthesis in both biochemical (upper panel) and graph (lower panel) ways of representations. The rst one
comes from KEGG database, the second one is obtained using Pajek software on Ma and Zeng [2] database. Table 1 allows for the conversion of the
upper to the lower panel. It is worth noting only green labelled enzymes are present in yeast. The lower panel includes some other peripherical arcs
and nodes linking fatty acid biosynthesis to the whole network.
4646
Table 1
The table reports the conversion key from the biochemical to graph
representations and viceversa
Node number
Metabolite name
4246
5745
5746
5747
5748
5749
5750
4620
5751
5752
5753
4619
5754
5755
5756
5757
5758
5223
5759
4688
5760
5761
5762
4633
4364
5763
5764
249
154
5272
2934
But-2-enoyl-[acp]
Butyryl-[acp]
3-Oxohexanoyl-[acp]
(R)-3-Hydroxyhexanoyl-[acp]
trans-Hex-2-enoyl-[acp]
Hexanoyl-[acp]
3-Oxooctanoyl-[acp]
(3R)-3-Hydroxyoctanoyl-[acyl-carrier protein]
trans-Oct-2-enoyl-[acp]
Octanoyl-[acp]
3-Oxodecanoyl-[acp]
(3R)-3-Hydroxydecanoyl-[acyl-carrier protein]
trans-Dec-2-enoyl-[acp]
Decanoyl-[acp]
3-Oxododecanoyl-[acp]
(R)-3-Hydroxydodecanoyl-[acp]
trans-Dodec-2-enoyl-[acp]
Dodecanoyl-[acyl-carrier protein]
3-Oxotetradecanoyl-[acp]
(3R)-3-Hydroxytetradecanoyl-[acyl-carrier protein]
trans-Tetradec-2-enoyl-[acp]
Tetradecanoyl-[acp]
3-Oxohexadecanoyl-[acp]
(3R)-3-Hydroxypalmitoyl-[acyl-carrier protein]
2-Hexadecenoyl-[acyl-carrier protein]
trans-Hexadec-2-enoyl-[acp]
Hexadecanoyl-[acp]
Palmitate
Palmitoyl-CoA
trans-Hexadec-2-enoyl-CoA
3-Dehydrosphinganine
The bolded elements are the ones included in the biochemical pathway.