In silico metabolic reconstruction of Fe-S cluster biogenesis in yeast

Rui Alves

Ciencies Mediques Basiques

Universitat de Lleida

04/21/23 2

Introduction

Novel pathways are being discovered with genome sequencing

Well known proteins are shown to be involved in some of the pathway but information about how the pathway structure is formed in unknown.

Finding these circuits & pathways is an important problem

Objective of the research line

Develop coherent framework where different computational methods and data sets are integrated to predict the connectivity of biological pathways & circuits. Today: focus on the biology and the

reconstruction of FeS cluster biogenesis in yeast

Fe-S clusters

Iron-Sulfur Clusters are coordinated ions that participate in electron transfer

Protein Cysteine

Protein Cysteine

Fe FeS

S

e- e-

04/21/23 5

What is known about FeSC biogenesis

About 15 different mitochondrial proteins are known to be involved in yeast

The assembly process is ill-understood

All 15 proteins have one thing in common

Phenotype of FeSC machinery deletion mutants

WT

Fe Level

WT

FeSC Dependent Protein Activity

Fe Accumulates FeSC dependent protein activity is impaired

FeSC biogenesis in a nutshell

Fe S

Scaffold Scaffold

FeSCSynthesis

Transfer

RepairHolo-P

Damaged FeSC

Apo-PHolo-P

FeSC

Scaffold Scaffold

(S)

(T)

(R)

04/21/23 8

The proteins and their function

Understanding the role of seven proteins in S. cerevisiae. FeS cluster biogenesis:

Grx5, Arh1-Yah1, Ssq1-Jac1-Mge1, Nfs1

Grx5 is involved in FeSC biogenesis in S. cerevisiae

Glutaredoxin Mediates glutathionylation state of Cys

residues May mediate protein-protein disulfide bridge

reduction (Belli et al. 2002, Tamarit et al. 2003, JBC)

FeSC coordinate (mostly) with Cys residues

Is Grx5 regulation of Cys reduction state in any specific protein(s) involved in FeSC biogenesis sufficient for phenotype?

04/21/23 10

Predicting partners for Grx5: the protocol

•Combine literature analysis, phylogenetic analysis of fully sequenced genomes and in silico protein docking to predict the most likely targets of Grx5

04/21/23 11

•Literature co-occurence of genes can be taken as a signal that they are functionaly related and maybe interact physically

•iHOP performes this type of analysis automatically

iHOP literature network reconstruction

04/21/23 12

Similarly, if Grx5 is absent in all genomes in which protein B is present there is a likelihood that they perform the same function!

Grx5 has highest CIs with the scaffold proteins

Using phylogenetic profiles to predict protein interactions

Sequence (Grx5) script

Database of profiles for each protein in each organism

Database of proteins in fully sequenced genomes

Protein id Grx5

Target Genome

Homologue in Genome 1?

Homologue in Genome 2?

…

Grx5

B

C

…

Y

N

Y

…

N

Y

N

…

…

…

…

…

Grx5 B

00i/number of genomes<1C

1j/number of genomes

Grx5 1

C 0.9

… …

B 0.11

… …Proteins (Grx5 and C) that are present and absent in the same set of genomes are likely to be involved in the same process and therefore interact

2

Calculate coincidence index

04/21/23 13

Low level study of docking interactions in silico

…SSQIE……SSQEE…

Sequence of known structure

OPTIMIZE

DOCK

THREAD

Homologue sequence for structure prediction

Scaffold proteins

Nfs1

(Cys Desulfurase)

Possible partners of Grx5 in FeSC biogenesis

Grx5

Nfs1

Isa2

Isa1

Isu1 Bibliography

Docking

Phylogeny+ Docking

FeSC biogenesis in a nutshell

Fe S

Scaffold Scaffold

FeSCSynthesis

Transfer

RepairHolo-P

Damaged FeSC

Apo-PHolo-P

FeSC

Scaffold Scaffold

(S)

(T)

(R)

04/21/23 16

Possible roles of Grx5 in FeSC biogenesis:regulation of glutathionylation

HSGlutathione

Grx5

S-SGPP

S

Possible roles of Grx5 in FeSC Biogenesis:Recovery of dead-end complex

Grx5

SHScaffold HS Nfs1

SScaffold S Nfs1

04/21/23 18

Studying the effect of Grx5: The modeling

Nfs1-SSG Nfs1

Grx5,…

11 1 21 5 1... ...

hg gd NfsNfs Grx Nfs

dt

g<0 inhibits flux

g=0 no influence on flux

g>0 activates flux

04/21/23 19

Studying the effect of Grx5: the protocol

•Create models for alternative networks

•Normalize equations and scan parameters

•Compare simulations with known systemic behavior to validate or invalidate alternatives

Model reproduces effect of gene deletion on protein activity if Grx5 recovers Nfs1 activity

Recovering Nfs1 and Scaffold

FeSC Dependent Protein Activity

Not recovering Nfs1 and Scaffold

Belli et al. MBC 13:1109

1000s of simulations

1

0.1

0.50.5

WT

1

0.1

Model reproduces effect of gene deletion on protein activity if Grx5 recovers Nfs1 activity

Fe Levels

WT Recovering Nfs1 and Scaffold

Not recovering Nfs1and Scaffold

Belli et al. MBC 13:1109

1 1

1000s of simulations

Grx5 is predicted to dock facing the Nfs1 active center

Alternative Grx5 Binding solutionsAlternative

Grx5 Binding solution

Nfs1 dimer

Active center Nfs1 Cys residue

Active center Grx5 Cys residue

Conclusions:

Grx5 modulates Nfs1 and Scaffold activity/Interactions

Possible Modes of action for Grx5

9Reproducing experimental phenotype?

No 6

Yes 3 Nfs1-Scaffold

Predictions for Grx5:

Grx5 modulates Cysteine Desulfurase (Nfs1) and scaffold activity and maybe the interaction between both

04/21/23 25

Grx5 interacts with scaffold in two-hybrid assay

Negative Controls Grx5 Scaffold

Positive Control

Arh1-Yah1 proteins are involved in FeSC biogenesis in S. cerevisiae

Arh1-Yah1 Ferredoxin Reductase – Ferredoxin

These proteins supply/drain electrons from other processes

What is their role in FeSC biogenesis?

04/21/23 27

Possible partners of Arh1-Yah1 in FeSC biogenesis

Isu2

Arh1

Isa2

Isa1

Docking + Phylogeny

Bibliography + Docking + Phylogeny

Yah1Yfh1

Nfu1

Isu1

Possible modes of Arh1-Yah1 action

Fe S

Scaffold Scaffold

FeSCSynthesis

Transfer

RepairHolo-P

Damaged FeSC

Apo-PHolo-P

FeSC

Scaffold Scaffold

Arh1/Yah1

Arh1/Yah1

Arh1/Yah1

(S)

(T)

(R)

Model reproduces gene deletion effect on FeSC dependent activity if Arh1-Yah1 act on

ST

T RS

R

RST RT

S

ST

FeS

C D

epen

dent

Pro

tein

Act

ivit

y

Li et al. JBC 276:1503Lange et al. PNAS 97:1050

1000s of simulations

WT

1

0.2

Model reproduces Fe accumulation upon gene deletion if Arh1-Yah1 act on ST

Fe Levels

Li et al. JBC 276:1503Lange et al. PNAS 97:1050

R RT

RST

S T

RS

S

ST

1000s of simulations

WT

1

Conclusions for Arh1-Yah1:

Arh1-Yah1 act on ST steps of FeSC biogenesis

Possible Modes of Arh1-Yah1 action

7Reproducing experimental phenotype?

No 5

Yes 2S

ST

Ssq1-Jac1-Mge1 proteins are involved in FeSC biogenesis in S. cerevisiae

Ssq1-Jac1-Mge1 Chaperone – Co-Chaperone – Nucleotide Release

Factor These proteins help fold-stabilize other proteins

It has been suggested that they help stabilize the FeSC in the scaffold for transfer.

Is this role necessary to justify the phenotypes?

Possible Modes of Chaperone Action

Fe S

Scaffold Scaffold

FeSCSynthesis

Transfer

RepairHolo-P

Damaged FeSC

Apo-PHolo-P

FeSC

Scaffold Scaffold

Ssq1/Jac1/Mge1

(S)

(T)

(R)

04/21/23 34

Model reproduces gene deletion effect on FeSC dependent activity if Arh1-Yah1 act on

3 steps

Stability Fold

FeS

C D

epen

dent

Pro

tein

Act

ivit

y

1000s of simulations

WT

1

0.2

Model reproduces Fe accumulation upon gene deletion if chaperones act on stability

Fe Levels

FoldStability

1000s of simulations

WT

1

Conclusions:

Scaffolds need to act only on folding in FeSC biogenesis

Possible Modes of Ssq1-Jac1-Mge1 action

3Reproducing experimental phenotype?

No 1

Yes 2Fold

Stability/Fold

Nfs1 is involved in FeSC biogenesis in S. cerevisiae

Nfs1 Cisteine Desulfurase

This protein provides sulfur for the biogenesis

It has been shown in vitro that it is able to repair FeSC in situ and bypass the biogenesis pathway

Is this role important?

Possible modes of Nfs1 action

Fe S

Scaffold Scaffold

FeSCSynthesis

Transfer

RepairHolo-P

Damaged FeSC

Apo-PHolo-P

FeSC

Scaffold Scaffold

Nfs1

Nfs1

(S)

(T)

(R)

Model reproduces gene deletion effect on FeSC dependent activity if Nfs1 acts on SR

R S

RS

FeSC Dependent Protein Activity

1000s of simulations

WT

1

0.2

Model reproduces Fe accumulation upon gene deletion if Nfs1 acts on S

Fe Levels

R S

RS

1000s of simulations

WT

1

Conclusions:

Nfs1 needs only to act on synthesis but can also act on repair of FeSC

Possible Modes of Nfs1 action

3Reproducing experimental phenotype?

No 1

Yes 2 S

SR

Summary

Arh1-Yah1 acts on Synthesis and transfer of FeSC

Grx5 modulates Cysteine Desulfurase (Nfs1) activity and maybe Scaffold activity

Ssq1-Jac1-Yah1 act on folding FeSC proteins

Nfs1 acts on in situ synthesis of clusters

Yfh1 does not modulate Fe import into the mitochondria

Alves et. al. 2004 Proteins 57:481Vilella et. al. 2004 Comp. Func. Genomics 5:328

Alves et. al. 2004 Proteins 56:354

Alves & Sorribas 2007 BMC Systems Biology

04/21/23 43

PS: The reconstruction method

04/21/23 44

Acknowledgments

Albert Sorribas Enric Herrero Armindo Salvador

FCT Spanish Government NIH (Mike Savageau)

04/21/23 45

Possible partners of Nfs1 in FeSC biogenesis

Nfs1

Isa2

Isa1

Docking

Docking + Phylogeny

Bibliography + Docking + Phylogeny

Isu1

Nfu1

Isu2

Metabolic Reconstruction: FeSC biogenesisThe view from here

Test the predictions

Extend the analysis to a variety of bacteria (Preliminary results for E. coli and Buchnera)

Create an interactive database/server to implement the methodology and apply it to other systems

Process of interest

Determining gene+proteins involved in process

Refining using automated literature analysis

Refining using phylogenetic and “omics” data

Protein Structure

(PDB/ Models)

In silico Protein Docking

Predict networks

Two Hybrid Screens

Predict networks

Co-evolution analysis

Predict networks

Omics data analysis

Predict networks

Baeysian network/human curation for alternative network structures

Automated creation of mathematical models

Analysis of model behavior

No model is validated by existing data

Some models are validated by comparison with existing data

Design new experiments to distinguish between alternatives

Reconstructing Metabolism and Investigating Design Principles in Molecular Biology II

Rui Alves

Ciencies Mediques Basiques

Universitat de Lleida

Outline Metabolic Network Reconstruction

Iron-Sulfur Cluster Biogenesis Pathway in S. cerevisiae.

Pathway Evolution Amino acid biosynthetic pathways protein

composition

Design Principles Regulatory Design in Networks

Two Component Systems• Mono-functionality vs. Bi-functionality of Sensor Proteins

Studying an organism

…ACTG…

>Dna

MAACTG…

>DNA Pol

MTC…

Stress

Measure Response

Find signatures for physiological dynamics in

genomic data

Stress Regulation Some genes are expressed specifically in response

to a stimulus.

For such genes, the amino acid composition of the proteins can be biased to facilitate their own synthesis and the physiological response.

This creates a signature of the response in the protein composition

Cognate Bias Cognate Bias

Carbon & Sulfur fixing enzymes have a lower content of amino acids with sidechains containing a lot of carbon or sulfur atoms (Baudouin-Cornu et al. Science 293:297)

Amino acid biosynthetic enzymes have a lower content of their cognate amino acid (Alves & Savageau 2005 Mol. Microbiol. 56:1017-34)

Regulation of amino acid biosynthetic pathways gene expression

aa rich medium t

Switch to AA poor medium

No cognate aa in biosynthetic proteins

Lots of cognate aa in biosynthetic proteins

aa levels

protein levels

aa poor medium

Effect of relative content of cognate amino acid on protein synthesis

Low cognate aa content in biosynthetic pathway aa

levelsProtein Synthesis

tHigh cognate aa content in biosynthetic pathway

aa levels

Protein Synthesis

t

The Low Cognate Bias Hypothesis

Thus, we hypothesize that evolution would lead to the selection of amino-acid biosynthetic enzymes that have a relatively low content of their cognate amino acid. We call this the “cognate-bias hypothesis”.

Test Cases: E. coli S. typhimurium B. subtilis

Calculating the bias

Protein Sequences

GeneBank

Specific genome

KEGG

Metacyc

Literature

aa-biosynthesis proteins

Proteome

Growing cells aa composition

Control Groups:

Calculate aa composition

Rank w/respect to proteome

Av.0.5

1

0Low

High

P(aa)

…

There is low cognate bias in aa biosynthesis

Positive correlation between relative cognate aa composition and specific activity of proteins

Specific activity

Specific activity

Protein copies

Protein copies

aa synthesis

aa synthesis

P<0.003

Negative correlation between relative cognate aa composition and number of proteins in pathway

Average aa content

+

++

+++

++++

1 2 3 4

P<0.003 (except Bs)

04/21/23 60

Higher bias is due to functional requirements

Active Center

04/21/23 61

Exceptions to the cognate bias hypothesis

Functional Reasons Active Centers (Phe, Tyr) Dimerization Domains (Asp)

Low cognate bias is present in organisms with the full complement of pathways

04/21/23 63

Environmental effects in cognate bias

The relative abundance of amino acid should influence the pressure to keep a low cognate amino acid content in biosynthetic pathways

E. coli (or S. typhimurium) and B. subtilis live in different environments

Amino acid availability is different

Is there a signal for these differences?

04/21/23 64

Differences in amino acid composition of distinct habitats

Low amounts Ala, Asp, Gly, Ser, Thr

High amounts Arg, Glu, Lys, Trp, Tyr

Low amounts Arg, Glu, Lys, Trp Tyr

Intermediate amounts

Ala, Asp, Thr

High amounts Gly, Ser

Savageau 1983 Am. Naturalist. 122:732

Soil Gut

04/21/23 65

Positive correlation between environmental levels of aa and cognate bias

Rank of Environmental aa

Rank of Cognate Bias

Presence of aa in environment decreases pressure for low cognate bias

P<0.004

04/21/23 66

Environmental availability influences cognate bias

Thus the effect of amino acid availability in the environment and the consequent regulation and physiological dynamics of gene expression does leaves a signature in the cognate bias of the different pathways.

Conclusions:

The low cognate bias hypothesis is supported by: Analysis of protein composition Analysis of protein specific activity Analysis of pathway length

Some factors that overcome selective pressure for low cognate bias Functional requirements for specific cognate residues Environmental factors

Alves & Savageau 2005 Mol. Microbiol. in press

Amino acid composition: the view from here

Extend the work for other bacteria and attempt to create organism/environment biosignatures

Analyze ribosomal proteins amino acid bias, amino acid transport proteins bias and catabolic protein bias

Analyze influence of oxydizability on selection of surface amino acids in proteins

Question Given that environment and bias are related, and

that aa production is mostly biotic, can different broad environments be identified by their fractional amino acid composition?

Environmental amino acid data

From the literature : Soil

10 different soil types, different times of the year Internal

Intestine of different animals Oceans, Rivers, lakes

13 sites with determinations all over the world and all year sampling

Looking for amino acid signaturesAla Cys Asp …

Pacific 0.15 0 0.07 …

Artic

Wilderness

.

.

.

0.14 0 0.08 …

Principal component analysis using the covariance matrix

Environments are well separated with three principal components

Questions Can different broad environments be identified by

their fractional amino acid composition?YES!

Similarly, can groups of organisms be classified using such a broad feature as their fractional amino acid composition?

Obtaining the biological data

Identify environment

KEGGTake genomes

Get protein sequences

Estimate the relative

composition of each cell type

KEGG

Literature

Different proteins contribute differently

Proteins have different levels of expression

Highly expressed protein will contribute more for the relative amino acid composition

Karlin (late 90s) proposed that highly expressed proteins can be identified by their codon bias

Rybossomal proteins are on average highly expressed

Using Codon Bias to weight gene expression

Calculate average codon frequency for rybosomal proteins

(CUAA1,…,CUAA20)

Calculate codon frequency for each protein

(CUAA1Pi,…,CUAA20Pi)

Normalize

Weight number of amino acids in protein j by Normalized DPj

63 2

, ,Pri=1

1,

2,

=

/ 1

1 /

k i Ribosome i oteink

k k

k k

CU CU

d Exp Max in genome Exp Exp Max in genome

d Max in genome

Do Principal Components Analysis

Segregation follows phylogeny

Archaea

Eukaryotes

Bacteria

Archaea

Eukaryotes

Gram PositiveGram Negative

Environmental segregation in Bacteria

Sea

Soil

Water

Parasite

Versatile

Environmental segregation seen in power spectrum

Gram PositiveGram

NegativeArchaea

Eukaryotes

SoilSea

WaterParasiteVersatil

e

Fou

rier

Tra

nsfo

rmF

ouri

er T

rans

form

Protein Length Protein Length

Relative content in proteins Relative content in proteins

Alanine Alanine

Questions Can different broad environments be identified by their

fractional amino acid composition? YES!

Similarly, can broad groups of organism be identified using such a broad feature as their fractional amino acid composition? To some extend yes. Broad phylogenetic grouping is

captured; environmental grouping is captured with less certainty, dependent on environment + phylogenetic group.

Final Question Environments can be identified by their aa

composition The cognate environment of a cell can be predicted

to some extent based on the cell’s amino acid content

Is there a correlation between the fractional amino acid composition of an environment and that of genomes from that environment?

Looking for co-adaptation

1, 1 1, 1 20, 1 20, 1, ,..., ,O Env O EnvAA AA AA AA

Calculate correlation coefficients

Final Question There is significant correlation between amino acid

content of cells and that of environment Signal is not strong enough to show adaptation to

specific environmental niches

Summary There are environmental and cellular signatures for

the amino acid composition There is correlation between the two types of

signatures but not enough signal for this correlation to be predictive.

Acknowledgments

Mike Savageau Mike Sternberg Albert Sorribas Enric Herrero Armindo Salvador

PGDBM JNICT FCT Spanish Government Portuguese Government NIH (Mike Savageau) DOD (ONR) (Mike Savageau)

Metabolic Reconstruction: the view from here (I)

The “Putting it together” Section

Continue FeSC work

Use Know how to reconstruct TCS network in M. xanthus

Analyze more TCS designs

Analyze aa biosyntesis and enzyme networks evolution

Grx5 interacts with Scaffold in Two-Hybrid assay

Two-hybrid analysis of the interaction between Grx5 and Scaffold. Numbers over bars indicate the beta-galactosidase activity (Miller units) in cultures of S. cerevisiae cells co-transformed with plasmids pGBT9 and pACT2 vectors alone, or derivatives expressing the respective Gal4 fusion proteins with Grx5, Scaffold, and two proteins known to interact. Results are the mean of three independent experiments.

Negative Controls Grx5 Scaffold

Positive Control

ATP binding domain important in functionality of sensor

Alves & Savageau 2003 Mol. Microbiol. 48:25

…SSQIE……SSQ-E…

Sensor with known structureSensor sequence for structure prediction

100s of structure predictions

Predicted to be Bifunctional Sensor

Predicted to be Monofunctional Sensor

Differences in ATP lid