Microbial Growth on Multiple Substrates

Professor Dhinakar S. Kompala Department of Chemical and Biological Engineering ni!ersit" of Colorado Boulder# Colorado $%&%'
(ith Pol"math programs written b" ).S. *ogler and Som Ghosh

Modifications to Monod Model

The empirically based Monod growth rate equation

has become popular, compared to the other proposed rate equations for cell growth, due to its similarity with the mechanistically derived

Michaelis-Menten rate law for enzyme-catalyzed reaction rate. The Monod equation is capable of explaining or simulating the exponential growth phase followed by the decelerating growth phase in the cell concentration during batch growth dynamics, when coupled with dynamic balance equation for the substrate concentration:

where YC/S is the stoichiometric yield coefficient of grams of cell mass produced from a gram of substrate.

igure !. Typical growth phases in batch cultures of bacterial cells

The initial lag and accelerating growth phases which are not simulated by the Monod growth rate equation can be easily simulated by ma"ing a slight modification to the original Monod equation given above. This modification invo"es a rate-limiting enzyme involved in the growth processes that may not be present at sufficient levels in the inoculum or starting culture. #hile cell growth is a complex process mediated by thousands of enzymes, it may be sufficient to hypothesize $ust a single enzyme that may be rate-limiting during the initial lag growth phase for the purpose of simulating the lag.

%ncorporating the effect of varying "ey enzyme concentration into the Monod growth "inetics, we can write

&'.(.).*+ where ER represents the relative amount of the "ey enzyme in the cell. This modified Monod rate equation follows the Michaelis-Menten rate equation closely, which is more correctly written as

&'.(.).)+ where ET represents the total enzyme concentration in the reaction mixture.

'elative "ey enzyme content inside the cells, ER, may be written as , where e is the intracellular content of the "ey enzyme, with the units of g enzyme,g cell mass, and emax is is the maximum enzyme concentration in the cell, also with units of g enzyme,g cell mass. The balance equation for the intracellular enzyme content can be written in terms of eCc, which has the units of g enzyme,culture volume, as

&'.(.).-+ where a and b are enzyme synthesis and degradation rate constants. .sing the product rule, we can expand the above equation to write the balance equation for e as:

&'.(.)./+ The last term in the above balance equation for the intracellular enzyme is the dilution term due to cell growth, which will be obtained similarly for all intracellular species. This equation can be solved to produce the curve added at the bottom of the batch growth dynamics as shown in igure 0. The maximum level of the intracellular enzyme emax can be determined easily by setting the above equation to zero and solving for its steady state value. %n terms of the model parameters, the maximum level of intracellular enzyme obtained during the exponential growth phase &when Cs is much greater than Ks+ can be derived as

&'(.).(+

1tarting with a low or zero content of the "ey enzyme needed for growth on a given substrate &representative of poor quality starter culture or inoculum+, the presence of the substrate induces the synthesis of the "ey enzyme. The low or zero enzyme content causes the initial lag phase of no cell growth while increasing enzyme content results in the accelerating growth phase. The enzyme concentration reaches and stays at its maximum, emax, during the exponential growth phase, which shows up as a linear increase in the logarithmic plot of cell concentration. The slope of this line on the semi-logarithmic plot below is the maximum specific growth rate, mmax . 2s the substrate gets depleted, the growth rate or the slope of this curve decelerates and becomes zero when the substrate is completely consumed. #ith the substrate no longer present, the "ey enzyme synthesis stops and enzyme degradation during the stationary and death phases reduces the intracellular enzyme content to low levels. P+,-M./) code0

Calculated !alues of the DE1 !ariables

3ariable initial value minimal value maximal value final value t 5c 5s !! 4 4.! ) 4 4.! 6.6*7-!! ) !.( ) ) !.( 6.6*7-

e 6.*7-4!.407-4) 8cs 4.)

6.*7-4-

!.4-7-4)

4.)

4.)

4.)

alpha !.47-4) beta 4.4mumax 4.9 :s

!.47-4)

!.47-4)

!.47-4)

4.4-

4.4-

4.4-

4.9

4.9

4.9

4.!

4.!

4.! !.4-*7-4)

4.! !.4-*7-

emax !.4-*7-4) 4) !.4-*7-4) 7' 4.(664.9/6/)9*

4.(66-

4.99(--4)

mu 4./90*)!(./967-!4 4.6/)0((9 (./967-!4

+DE 2eport 32K*456 ;ifferential equations as entered by the user <!= d&5c+,d&t+ > mu ? 5c <0= d&5s+,d&t+ > -!,8cs ? mu ? 5c <*= d&e+,d&t+ > alpha ? 5s,&:s @ 5s+ - beta?e mu?e

7xplicit equations as entered by the user <!= 8cs > .) <0= alpha > .444! <*= beta > .4<)= mumax > .9 <-= :s > .! </= emax > alpha,&mumax@beta+ <(= 7' > e,emax

<6= mu > mumax ? 7'?5s,&:s @ 5s+

Figure 2. Dynamic profiles of substrate concentration and intracellular enzyme content along with the logarithm of cell mass concentration.
5lic" here for the Aolymath 5ode

7.4.&.8 Se9uential Growth : Diau;ie

5onsider the cell growth medium in a batch bioreactor that contains two different carbon substrates, 1! and 10, which are each capable of supporting cell growth. or the growth on the substrate !: 5ells @ 1ubstrate ! &e.g. glucose+ B more 5ells @ products, the modified Monod growth rate expression is

&'.(.).6+ or growth on the substrate 0:

5ells @ 1ubstrate 0 &e.g. lactose+ B more 5ells @ products, the modified Monod growth rate expression is

&'.(.).9+ where m1 and m2 are the specific cell growth rates on individual substrates 1! and 10 respectively.

#hen both the substrates are present in a batch bioreactor, microbial cells do not consume both substrates simultaneously or additively but instead grow on the sugars sequentially. That is, the maximum specific growth rate on mixed substrates: . The sequential consumption results in an interesting pattern of the cells first consuming one preferred substrate and then after an intermediate lag phase, consuming the remaining less preferred substrate. That is, the less preferred substrate is not utilized until the preferred substrate is completely consumed and is no longer available in the growth medium. This sequential utilization of two substrates in batch cultures has been observed in numerous experiments by Monod, who has termed this phenomenon as the CdiauxieC &Dree" for two growth phases+.

The consistent characteristic of the diauxic growth phenomenon is that the preferred substrate provides the faster growth rate, i.e. mmax,! E mmax0. The emzyme growth curve along with the cell and substrate concentration are shown in igure *.

Figure 3. Diauxic growth of bacteria on two substrates, showing also the intracellular content of the two hypothetical key enzymes for consuming each substrate.

The diauxic growth is observed in most or all welldocumented batch cultures on multiple substrates. The preference for faster growth rate in the first growth phase has been suggested to be a consequence of evolutionary pressures on the microbes to grow at the fastest growth rate possible <:ompala et al., !96/=. %f the cells were to grow simultaneously on both substrates it has been suggested that the cell growth rate will be reduced to an average of the two growth rates, rather than an additive growth rate.

The diauxic growth phenomenon has been modeled with the modified Monod equations discussed above, and the evolutionary ob$ective of maximizing the instantaneous growth rate accomplished through hypothesized CcyberneticC variables. The cybernetic variables ui and vi represent the outcome of intracellular regulatory processes controlling enzyme synthesis and activity respectively, and are determined through maximization of instantaneous growth rate.

&'.(.).!4+

&'.(.).!!+

&'.(.).!0+

for where

i > !,0

&'.(.).!*+

&'.(.).!)+

e.g.

and

The last term in the equation &'.(.).!*+ is the typical dilution of an intracellular species due to cell growth. ;uring the exponential growth on high concentration of a single substrate, the maximum level of the intracellular "ey enzyme can be shown as

&'.(.).!-+ #ith these CcyberneticC model equations, the diauxic growth phases shown above have been simulated with typical model parameters values.

The model parameters are the Monod growth parameters for growth on each single substrate &which can be determined experimentally+ and additional model parameters ai and bifor the synthesis and degradation of the two "ey enzymes. 2s the rate-limiting "ey enzymes are hypothetical, even though they may be identifiable in many cases &such as, b-galactosidase for the utilization of lactose+, these additional model parameters are given some representative values, rather than experimentally measured.

E;ample 2.7.4.8.80 Diau;ic Growth0 Ma;imi<ation of instantaneous growth rate

The cybernetic variables ui and vi represent the optimal response of the enzymatic synthesis and activity to maximize the instantaneous growth of the cells for any given medium composition. Therefore, even if the substrate ! is present at lower concentration than the substrate 0 at the beginning of batch growth on mixed substrates, as long as the instantaneous growth rate m1 is greater than the rate m2, more of the enzymes need for the utilization of substrate ! will be synthesized and active &u1 E u2 and v1 v2 +. This is illustrated in the following numerical simulation, using the initial concentrations of ) g,F for substrate ! and 04 g,F for substrate 0. Gowever, because of the model parameter values used &mmax,! > 4.9 hr-!, mmax,0 > 4./ hr-!, :! > 4.! g,F, :0 > 4.- g,F, ai > 4.444! hr! , bi > 4.4- hr-!+ and initial conditions for the two "ey enzymes &e!,4 at closer to its maximum value of 6.* x !4-- and e0,4 at a lower value of ! x !4-/+, the growth rate on substrate ! m! remains higher than m0 until substrate ! is completely consumed. 5onsequently cybernetic variables u1 and v1 for the first enzymeHs synthesis and activity remain close to their maximum value of ! during the first growth phase. P+,-M./) code0 32un this program with S/=** +DE algorithm6

Calculated !alues of the DE1 !ariables

3ariable initial value minimal value maximal value final value t 5c 4 4.! 4 4.! 0.)4/7-0!4 9.( ) !4 9.(

5s! ) 0.)4/7-05s0 04 *.-!*7-0e! 6.*7-4!.*0*7-4e0 !.47-4/ !.*-*7-4) mu!max 4.9 mu0max 4./ :! :0 beta! 4.4beta0 4.4alpha! !.47-4) alpha0 !.47-4) 4.! 4.4.44.9

-0.4(07-!!

04

!.*0*7-4-

!.4)97-4)

0.*697-4(

!.-!7-4)

4.9

4.9

4./

4./

4./

4.! 4.4.4-

4.! 4.4.4-

4.! 4.-

4.4-

4.4-

4.4-

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

e!max !.4-*7-4) 4) !.4-*7-4) e0max !.-*67-4) 4) !.-*67-4)

!.4-*7-4)

!.4-*7-

!.-*67-4)

!.-*67-

7! 4.(664.!0-6-*70 4.44/4.664/(

4.!0-6-*-

4.99/(*)0

4.44!--06

4.96!0*9!

mu! 4./90*)!/.!*/7-04.6/*!(0! /.!*/7-0mu0 4.44*64)9 -0.**97-!! 4.-)()!)0 /.90*7-0u! 4.99)-*)) -!(.(4090! 4.996(69/ 4.)/96((9 u0 4.44-)/-/ 4.44!0!4) !6.(4090! 4.-*4!00! mumax 4./90*)!/.90*7-04.6/*!(0! /.90*7-0v! ! 4.66/*-(6 8cs! 8cs0 v0 ! 4.) 4.) 4.44-)9-( ).4967-4) !

4.) 4.)

4.) 4.) !

4.) 4.)

-!.4-/)6(9

+DE 2eport 3S/=**6

;ifferential equations as entered by the user <!= d&5c+,d&t+ > &mu!?v!@mu0?v0+?5c <0= d&5s!+,d&t+ > -mu!?v!?5c,8cs! <*= d&5s0+,d&t+ > -mu0?v0?5c,8cs0 <)= d&e!+,d&t+ > alpha!?5s!,&:!@5s!+?u! beta!?e! - &mu!?v!@mu0?v0+?e!

<-= d&e0+,d&t+ > alpha0?5s0,&:0@5s0+?u0 beta0?e0 - &mu!?v!@mu0?v0+?e0

7xplicit equations as entered by the user <!= mu!max > .9 <0= mu0max > ./ <*= :! > .! <)= :0 > .<-= beta! > .4</= beta0 > .4<(= alpha! > .444! <6= alpha0 > .444! <9= e!max > alpha!,& mu!max@ beta!+ <!4= e0max > alpha0,& mu0max@ beta0+ <!!= 7! > e!,e!max <!0= 70 > e0,e0max <!*= mu! > mu!max?7!?5s!,&:!@5s!+ <!)= mu0 > mu0max?70?5s0,&:0@5s0+ <!-= u! > mu!,&mu!@mu0+ <!/= u0 > mu0,&mu!@mu0+ <!(= mumax > if &mu!Emu0+ then &mu!+ else &mu0+ <!6= v! > mu!, mumax <!9= 8cs! > .) <04= 8cs0 > .) <0!= v0 > mu0,mumax

5lic" here for the Aolymath 5ode

Figure E .!.".#$ %imulations of the cybernetic model, showing the profiles for the two substrate concentrations &%# and %2 on the left axis' and logarithm of cell mass &ln ( on the right axis' during a typical diauxic growth.

The intracellular enzyme levels for the same simulation are shown below to highlight the role of the cybernetic variable u1 and u2 in the synthesis of e1 and e2. These are plotted using the last program, by plotting e!,e!, u! and u0 vs time. P+,-M./) /.B,E e! 6.*7-4!.*0*7-4!.*0*7-4!.4)97-4)

e0 !.47-4/ !.*-*7-4)

0.*697-4(

!.-!7-4)

u! 4.99)-*)) -!(.(4090! 4.996(69/ 4.)/96((9 u0 4.44-)/-/ 4.44!0!4) !6.(4090! 4.-*4!00!

Figure E .!.".2$ %imulations of the cybernetic model showing the profiles of the two cybernetic )ariables &u# and u2 on the left axis' and two enzymes &e# and e2 on the right axis'

;uring the first growth phase, the cybernetic variable u1 ta"es values close to unity, indicating preferential synthesis of enzyme ! and repression &suppression of synthesis+ of enzyme 0 as u2 is close to zero. 2fter the first substrate is mostly consumed, the growth rate m1 on that substrate falls to zero, triggering the switch in the cybernetic variables and inducing synthesis of enzyme 0.

E;ample 2.7.4.8.>. 0 Diau;ic Growth0 Effect of Preculturing in Substrate >

%t may be suspected from the above simulation that preferential utilization of substrate ! is due to the high level of enzyme ! assumed as its initial value. This high initial value for e1 is chosen to indicate preculturing the inoculum in substrate !. %f the inoculum is precultured in substrate 0, the initial value for enzyme 0 should be higher and that for enzyme ! should be assumed much lower.

The two figures below show simulation results with the altered initial conditions for the two enzymes, while "eeping all other initial values and model parameter values identical to those in the above example. The diauxic lag gets significantly shortened, along with significant consumption of 10 during the first growth phase. Ievertheless, substrate ! is gradually preferred with increasing culture time and is completely consumed during the first growth phase. P+,-M./) code0 Calculated !alues of the DE1 !ariables

3ariable initial value minimal value maximal value final value t 5c 4 4.! 4 4.! !.((-7-!!4 -.( ) !4 -.(

5s! ) !.((-7-!5s0 !4 !.)!(7-!)

!.)!(7-!)

!4

e! !.47-4/ *.49/7-4e0 !.47-4) !.!/7-4) mu!max 4.9 mu0max 4./ :! :0 beta! 4.4beta0 4.4alpha! !.47-4) alpha0 !.47-4) 4.! 4.4.44.9

!.47-4/

/.)!07-4-

(.06)7-4-

!.)!7-4)

4.9

4.9

4./

4./

4./

4.! 4.4.4-

4.! 4.4.4-

4.! 4.-

4.4-

4.4-

4.4-

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

e!max !.4-*7-4) 4) !.4-*7-4) e0max !.-*67-4) 4) !.-*67-4) 7! 4.4494.09)*-9) 70 4./4.(-)6-

!.4-*7-4)

!.4-*7-

!.-*67-4)

!.-*67-

4.449-

4./49!(9/

4.)(*)-*!

4.9!/-(9(

mu! 4.446*)!/.0)67-!4.)(60*/6 /.0)67-!mu0 4.*(!)06/ !./*/7-!) 4.-00-(49 !./*/7-!) u! 4.40!9/)4.4!*)/!* 4./*-(09 4.0(/*(9*

u0 4.9(64*-4.*/)0(! 4.96/-*6( 4.(0*/04( mumax 4.-00-(49 4.*(!)06/ !./*/7-!) !./*/7-!)

v! 4.400)-(6 4.*6!9*9) 8cs! 8cs0 v0 ! 4.) 4.) !

4.4!*/)-

4.) 4.) 4.-(099()

4.) 4.) !

4.) 4.)

+DE 2eport 3S/=**6

;ifferential equations as entered by the user <!= d&5c+,d&t+ > &mu!?v!@mu0?v0+?5c <0= d&5s!+,d&t+ > -mu!?v!?5c,8cs! <*= d&5s0+,d&t+ > -mu0?v0?5c,8cs0 <)= d&e!+,d&t+ > alpha!?5s!,&:!@5s!+?u! beta!?e! - &mu!?v!@mu0?v0+?e! <-= d&e0+,d&t+ > alpha0?5s0,&:0@5s0+?u0 beta0?e0 - &mu!?v!@mu0?v0+?e0

7xplicit equations as entered by the user <!= mu!max > .9 <0= mu0max > ./ <*= :! > .! <)= :0 > .<-= beta! > .4-

</= beta0 > .4<(= alpha! > .444! <6= alpha0 > .444! <9= e!max > alpha!,& mu!max@ beta!+ <!4= e0max > alpha0,& mu0max@ beta0+ <!!= 7! > e!,e!max <!0= 70 > e0,e0max <!*= mu! > mu!max?7!?5s!,&:!@5s!+ <!)= mu0 > mu0max?70?5s0,&:0@5s0+ <!-= u! > mu!,&mu!@mu0+ <!/= u0 > mu0,&mu!@mu0+ <!(= mumax > if &mu!Emu0+ then &mu!+ else &mu0+ <!6= v! > mu!, mumax <!9= 8cs! > .) <04= 8cs0 > .) <0!= v0 > mu0,mumax

Figure E .!.".3$ %imulations of cybernetic model with altered initial conditions for the two enzyme le)els, with e2 * e#, reflecting the preculturing of inoculum on %2.
The gradual increase in the slope of semi log plot of cell concentration during the first growth phase is due to the gradual preference of substrate !, even though the inoculum is precultured on substrate 0 and has the enzyme 0 already available for its continued consumption. ;uring the later parts of first growth phase, more of the enzyme ! is synthesized &as is seen in the next simulation graph+ resulting in the rapid consumption of substrate ! and increasing growth rate &slope of the semi log curve+. 1ignificant availability of enzyme 0 at the end of first growth phase results in the reduced or non-existent diauxic lag phase before the second growth phase on the remaining substrate 0. P+,-M./) /.B,E0 e! !.47-4/ *.49/7-4e0 !.47-4) !.!/7-4) !.47-4/ /.)!07-4-

(.06)7-4-

!.)!7-4)

u! 4.40!9/)4.0(/*(9* u0 4.9(64*-4.(0*/04(

4.4!*)/!*

4./*-(09

4.*/)0(!

4.96/-*6(

Figure E .!."." %imulations of the cybernetic model showing the two cybernetic )ariables u1 and u2 along with the profiles of intracellular enzyme contents for the two key enzymes e1 and e2. +reculturing causes the initial )alue for e2 to be much higher than that for e1. E)en with the )alues, the model predicts an increasing preference for the substrate # during the first growth phase.

2.7.4.> Simultaneous tili<ation in Continuous Cultures

or continuous or chemostat culture of microbial growth on multiple substrates, the batch culture

balance equations given above can be modified to include the inlet and outlet terms as below:

&'.(.).!/+

&'.(.).!(+

&'.(.).!6+

!o"

i > !,0

&'.(.).!9+

where ; is the dilution rate, 51!,4 and 510,4 are the inlet concentrations of substrates 1! and 10 respectively. %n equation '.(.).!9 for intracellular enzymes, an additional synthesis rate constant a? is included to ensure a low level presence of each enzyme even in the absence of its substrate.

#ith these new dynamic balance equations for continuous cultures, the cybernetic model predicts the simultaneous utilization of both substrates at steady state for low dilution rates, as observed experimentally. 2t increasing dilution rates, the simultaneous utilization of both substrates changes gradually to #"e!e"ential utilization of the preferred &faster growth supporting+ substrate. 2t even higher &than the maximum growth rate possible in

the chemostat: + dilution rates, washout of cells from the chemostat is observed as also observed for single substrate chemostats.

E;ample 2.7.4.>.8 Simultaneous utili<ation of multiple substrates in continuous cultures

.sing the same model parameter values used in the previous examples &mmax,! > 4.9 hr-!, mmax,0 > 4./ hr-!, :! > 4.! g,F, :0 > 4.- g,F, ai > 4.444! hr-!, bi > 4.4- hr-! and the new parameter ai?> 4.4!a+, the above modified cybernetic model equations of continuous cultures can be simulated to plot the steady state concentration of cell mass, substrates ! and 0 over a range of dilution rates. The initial values for the different concentrations are immaterial &as long as cell mass concentration is not started at zero since there will be no spontaneous generation of life in a sterile bioreactor+ if we simulate the dynamic balance equations &'.(.).!/-'.(.).!9+ long enough for them to reach steady state. The inlet concentrations of the two substrates are chosen as !4 g,F each and the inlet nutrient feed is assumed to be sterile &i.e. cell mass concentration in the feed is zero+. P+,-M./) code0 3?ote0 Sol!e this program using the S/=** algorithm in P+,-M./)6

+DE 2eport 3S/=**6

;ifferential equations as entered by the user <!= d&5c+,d&t+ > &mu!?v!@mu0?v0+?5c - ;?5c <0= d&5s!+,d&t+ > -mu!?v!?5c,8cs! @ ;?&5s!45s!+

<*= d&5s0+,d&t+ > -mu0?v0?5c,8cs0 @ ;? &5s04 5s0+ <)= d&e!+,d&t+ > alpha!?5s!,&:!@5s!+?u! beta!?e! - &mu!?v!@mu0?v0+?e! @ alphastar <-= d&e0+,d&t+ > alpha0?5s0,&:0@5s0+?u0 beta0?e0 - &mu!?v!@mu0?v0+?e0 @ alphastar

7xplicit equations as entered by the user <!= mu!max > .9 <0= mu0max > ./ <*= :! > .! <)= :0> .<-= beta! > .4</= beta0 > .4<(= alpha! > .444! <6= alpha0 > .444! <9= e!max > alpha!,& mu!max@ beta!+ <!4= e0max > alpha0,& mu0max@ beta0+ <!!= 7! > e!,e!max <!0= 70 > e0,e0max <!*= mu! > mu!max?7!?5s!,&:!@5s!+ <!)= mu0 > mu0max?70?5s0,&:0@5s0+ <!-= u! > mu!,&mu!@mu0+ <!/= u0 > mu0,&mu!@mu0+ <!(= mumax > if &mu!Emu0+ then &mu!+ else &mu0+ <!6= v! > mu!, mumax

<!9= 8cs! > .) <04= 8cs0 > .) <0!= v0 > mu0,mumax <00= alphastar > .4!?alpha! <0*= 5s!4 > ) <0)= 5s04 > 04 <0-= ; > .)

#e run this program with ; > .! to ; > .9 with intervals of .! and collect the final values &equilibrium values+ and plot them against the corresponding ; values. Diven are the tables for ; > .), .- and ./.

Calculated !alues of the DE1 !ariables for D @ .4

3ariable initial value minimal value maximal value final value t 4 4 4.! !4 !4

5c 4.! !.-6494-* 5s! ) 4.49469)* 5s0 04 !9.9/!)0! e! 6.*7-49./9)7-4e0 !.47-4/ !.6/-7-4-

!.-6494-*

4.46!/(-!

!9.9/!)0!

04

6.*7-4-

!.4-*7-4)

!.47-4/

!.6/-7-4-

mu!max 4.9 mu0max 4./ :! :0 beta! 4.4beta0 4.4alpha! !.47-4) alpha0 !.47-4) 4.! 4.-

4.9

4.9

4.9

4./

4./

4./

4.! 4.4.4-

4.! 4.4.4-

4.! 4.-

4.4-

4.4-

4.4-

4.4-

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

e!max !.4-*7-4) 4) !.4-*7-4) e0max !.-*67-4) 4) !.-*67-4) 7! 4.(664.90!(-** 70 4.44/4.!!990/9

!.4-*7-4)

!.4-*7-

!.-*67-4)

!.-*67-

4.(66-

!.44444(0

4.44/-

4.!!990/9

mu! 4./90*)!4.*9)(/-4.6(006/( 4.*9)(/-mu0 4.44*64)9 4.44*64)9 4.4(4!9(9 4.4(4!9(9 u! 4.99)-*)) 4.6)9404.99)-*)) 4.6)940u0 4.44-)/-/ 4.44-)/-/ 4.!-49(4.!-49(mumax 4.6(006/( v! ! 4./90*)!4.*9)(/-! 4.*9)(/--

8cs! 8cs0

4.) 4.)

4.) 4.)

4.) 4.)

4.) 4.)

v0 4.44-)9-( 4.44-)9-( 4.!((60!/ 4.!((60!/ alphastar !.47-4/ 5s!4 5s04 ; !.47-4/ !.47-4/ !.47-4/

) 04 4.)

) 04 4.)

) 04 4.)

) 04 4.)

Calculated !alues of the DE1 !ariables for D @ .5 3ariable initial value minimal value maximal value final value t 4 4 4.! !4 !4

5c 4.! !.-)699(9 5s! ) 4.!*!6(99 5s0 04 !9.99(*! e! 6.*7-4!.40/7-4) e0 !.47-4/ -./407-4/ mu!max 4.9 mu0max 4./ :! :0 4.! 4.4.9

!.-)99-((

4.!094-4/

!9.99(*!

04

6.*7-4-

!.4-*7-4)

!.47-4/

-./407-4/

4.9

4.9

4./

4./

4./

4.! 4.-

4.! 4.-

4.! 4.-

beta! 4.4beta0 4.4-

4.4-

4.4-

4.4-

4.4-

4.4-

4.4-

alpha! !.47-4) !.47-4) alpha0 !.47-4) !.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

e!max !.4-*7-4) 4) !.4-*7-4) e0max !.-*67-4) 4) !.-*67-4) 7! 4.(664.9()//0/ 70 4.44/4.4*/!46/

!.4-*7-4)

!.4-*7-

!.-*67-4)

!.-*67-

4.(66-

!.444!0)(

4.44/-

4.4*/!46/

mu! 4./90*)!4.)96(6/ 4.6(*/!0) 4.)96(99( mu0 4.44*64)9 4.44*64)9 4.40!!*/( 4.40!!*/( u! 4.99)-*)) 4.9-9*)(4.99)-*)) 4.9-9*)(u0 4.44-)/-/ 4.44-)/-/ 4.4)4/-04.4)4/-0mumax 4.6(*/!0) v! 8cs! 8cs0 ! 4.) 4.) 4./90*)!4.)96(99( ! 4.) 4.) 4.)96(6/

! 4.) 4.)

! 4.) 4.)

v0 4.44-)9-( 4.44-)9-( 4.4)0*(-! 4.4)0*(-! alphastar !.47-4/ !.47-4/ !.47-4/ !.47-4/

5s!4 5s04 ;

) 04 4.-

) 04 4.-

) 04 4.-

) 04 4.-

Calculated !alues of the DE1 !ariables for D @ .A Calculated !alues of the DE1 !ariables

3ariable initial value minimal value maximal value final value t 4 4 4.! !4 !4

5c 4.! !.0(/-)!* 5s! ) 4.649--46 5s0 04 !9.999(!/ e! 6.*7-4!.4-!7-4) e0 !.47-4/ 0.!/*7-4/ mu!max 4.9 mu0max 4./ :! :0 beta! 4.4beta0 4.44.! 4.4.44.9

!.0(/-)!*

4.649--46

!9.999(!/

04

6.*7-4-

!.4-*7-4)

!.47-4/

0.!/*7-4/

4.9

4.9

4./

4./

4./

4.! 4.4.4-

4.! 4.4.4-

4.! 4.-

4.4-

4.4-

4.4-

alpha! !.47-4) alpha0 !.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

!.47-4)

e!max !.4-*7-4) 4) !.4-*7-4) e0max !.-*67-4) 4) !.-*67-4) 7! 4.(664.996!9/! 70 4.44/4.4!)4)!(

!.4-*7-4)

!.4-*7-

!.-*67-4)

!.-*67-

4.(66-

!.44404)!

4.44/-

4.4!)4)!(

mu! 4./90*)!4./90*)!4.6()(/9* 4.6446!-6 mu0 4.44*64)9 4.44*64)9 4.4460!94.4460!9u! 4.99)-*)) 4.9696)4* 4.99)-*)) 4.9696)4* u0 4.44-)/-/ 4.44-)/-/ 4.4!4!-9( 4.4!4!-9( mumax 4.6()(/9* v! 8cs! 8cs0 ! 4.) 4.) 4./90*)!4.6446!-6 ! 4.) 4.) 4./90*)!-

! 4.) 4.)

! 4.) 4.)

v0 4.44-)9-( 4.44-)9-( 4.4!40/) 4.4!40/) alphastar !.47-4/ 5s!4 5s04 ; !.47-4/ !.47-4/ !.47-4/

) 04 4./

) 04 4./

) 04 4./

) 04 4./

5lic" here for the Aolymath 5ode

Figure E .!."., %teady state )alues from the simulation of cybernetic model e-uations. .oth substrates are simultaneously utilized at low dilution rates. /t increasing dilution rates, substrate 2 is gradually re0ected in fa)or of substrate #. /t higher dilution rates, washout occurs with the steady state )alues same as the inlet concentrations.

%n mar"ed contrast to the batch culture results of sequential utilization of the two substrates, the continuous culture simulations &and the experimental data+ show the simultaneous utilization of both the substrates at low dilution rates. 2t increasing dilution rates, the second &less preferred or lower growth rate supporting+ substrate is not consumed completely or is re$ected in favor of the preferred &i.e. faster growth rate supporting+ substrate. 2t much higher growth rates, the washout steady state is observed with the two substrates and the cell mass reaching a steady state that is the same as their inlet concentration.

2.7.4.& Multiple Metabolic Pathwa"s in -east0

The brewerHs or ba"erHs yeast, Saccha"omyces ce"evisiae, presents an interesting example of the cybernetic ob$ective i.e. maximization of the cell growth rate through preferential utilization of a substrate or in this case a metabolic pathway over the others. The yeast cells have different pathways for consuming glucose:

$1% &lucose !e"mentative #ath'ay, which may be represented by the overall chemical equation, if glucose consumption for cell growth is ignored:

& .!.".21'

The Monod growth parameters for this growth process are:

(2) Ethanol oxidative pathway, with its o)erall chemical reaction &ignoring cell growth'$

& .!.".2#'

can also occur if ethanol and oxygen are both present in the culture medium.

The Monod growth parameters for this third growth process are:

and $(% &lucose oxidative #ath'ay, which is of course possible only in the presence of oxygen, again ignoring the glucose consumption for cell growth, the overall chemical reaction of this pathway can be represented as:

& .!.".22'

The Monod growth parameters for this second growth process are:

)y#othetical *uestion: %n a typical brewing experiment, if glucose and oxygen are both present in the culture medium, are both pathways used simultaneously or is one pathway preferentially utilized by the yeast, and if latter, which pathway is preferredJ

Iumerous brewers routinely ferment glucose to ethanol using yeast cells, without ta"ing any special precautions to eliminate oxygen from the culture medium. These fermentations are successful &in producing ethanol+ because cells preferentially use the !aste" fermentative pathway and do not produce the enzymes needed for slo'e" oxidative consumption of glucose even if oxygen is present in culture medium, until almost all the glucose has been fermented to ethanol. 2fter all glucose is fermented, it will be necessary to stop the batch fermentation to avoid the oxidative consumption of ethanol, which will occur in a subsequent or diauxic growth phase if oxygen is present.

The cybernetic model equations introduced earlier predicts the diauxic growth of yeast on glucose and ethanol in aerobic cultures, with small modifications to incorporate the specific case of ethanol generation from the fermentative pathway. The further modified cybernetic model equations for the yeast growth metabolism &to include the dynamics of intracellular storage carbohydrates, trehalose and glycogen, represented as 5T+ from Kones and :ompala &!999+ are given below for both batch &; > 4+ and continuous cultures.

&'.(.).0*+

&'.(.).0)+

&'.(.).0-+

&'.(.).0/+

&'.(.).0(+

&'.(.).06+

2he modified 3onod growth rates along the indi)idual metabolic pathways are$

&'.(.).09+

&'.(.).*4+

&'.(.).*!+

5ybernetic variables ui and vi are determined as before:

&'.(.).*0+ The symbols ! and +represent the stoichiometric constants for the production of ethanol, consumption of oxygen and the storage carbohydrates respectively.

The cybernetic model for yeast metabolism predicts the diauxic growth phases in the aerobic growth of yeast on glucose in batch cultures &with , > 4 and high values of -.a+. The model parameters were chosen to fit the experimental data from von Meyenburg &!9/9+ and are partially listed earlier with discussions on the three metabolic pathways.. P+,-M./) code Calculated !alues of the DE1 !ariables

3ariable initial value minimal value maximal value final value t 4 4 4.! 04 04

5c 4.! 0//9.*(6) 5g ) -!./-7@4) 5e 04 (94-.0()* 5o 4.4.4/*/)* e! 6.*7-44.*4/)0)) e0 !.47-4/ 4.!4-4!!9

0//9.*(6)

-!./-7@4)

04

(94-.0()*

4.4/*/)*

4.-

6.*7-4-

4.*/*!)6(

!.47-4/

4.!4(949*

e* !.47-4) 4.46!))!) 5t 4 -.(697-4( beta! gam* 4.6 mu!max 4.)) :o0 4.4! beta0 beta* mu0max 4.!9 mu*max 4.*/ alpha* :o* alpha! alpha0 4.( 4.6 4

!.47-4)

4.!0)()*)

4.!0!449)

4.( 4.6

4.( 4.6

4.(

4.))

4.))

4.))

4.4!

4.4!

4.4!

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.(

4.*/

4.*/

4.*/

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.* 4.06*4!69

4.* 0.0 4.* 4.*

e*max 4.06*4!69 4.06*4!69 4.06*4!69 e!max 4.0/*!-(9 4.0/*!-(9 4.0/*!-(9 e0max 4.**(4(6( 4.**(4(6( 4.**(4(6(

4.0/*!-(9

4.**(4(6(

7* *.-**7-4) *.-**7-4) 4.))4-)64.06((-9/ 7! *.!-)7-4) *.!-)7-4) !.*64)/() !.!/))!06

70 0.9/(7-4/ 0.9/(7-4/ 4.*04!*46 4.*!!-*-* :! 4.4:0 4.4! :* 4.44! 4.44.44.4-

4.4!

4.4!

4.4!

4.44!

4.44!

4.44!

mu! !.*(!7-4) !.*(!7-4) !.!964-66 4.-!0*)*0 mu0 -.-0*7-4( 4.4--0/6! 4.4-!!-) -.-0*7-4(

mu* !.!((7-4) -!.!00/696 4.*4))/0* -4.44!*9-0 mumax !.!964-66 !.*(!7-4) 4.-!0*)*0 !.*(!7-4)

u! 4.-*/(*/ 4.-0/4*09 9.64!()-( 4.9!!)(() u0 4.440!/09 4.440!/09 4.*6**(9( 4.49!44)6 8! 4.!/ 80 4.(8* 4.!/ 4.!/ 4.!/

4.(-

4.(-

4.(-

4./

4./

4./

4./

u* 4.)/!!4!! -9.!6-!0-) 4.)/!!4!! -4.440)600 alphastar 4.! 5g4 5s04 ; 4 4.! 4.! 4.!

!4 04 4

!4 04 4

!4 04 4

!4 04

v* 4.6-946*( -4.9*(4946 4.6-946*( -4.440(0** v! ! ! ! !

v0 4.44)4096 4.44)4096 4.!*-9)*! 4.4996)*0 sigmamuv 0.0-!9-) 0.*607-4) 4.-!()-)) 0.*607-4)

d5c 0.*607-40.*607-4!*6!.06!!*6!.06!gam! !4 gam0 !4 !4 !4 !4

!4

!4

!4

d5t 6.4907-4-4.66)9!4) 4.!!)*/(- -0.---7-4( phi! 4.)6 phi0 phi* phi) 4.)6 4.)6 4.)6

0 ! 4.9-

0 ! 4.9-

0 ! 4.9-

0 ! 4.9-

"la !444 5ostar

!444

!444

!444

4.!

4.!

4.!

4.!

+DE 2eport 32K*456

;ifferential equations as entered by the user <!= d&5c+,d&t+ > d5c <0= d&5g+,d&t+ > ;?&5g4 - 5g+ - &mu!?v!,8!@ mu*?v*,8*+?5c - phi)?&5t?d5c @ 5c?d5t+

<*= d&5e+,d&t+ > -;?5e @ & phi!?mu!?v!,8! mu0?v0,80+ ? 5c <)= d&5o+,d&t+ > "la ? &5ostar - 5o+ &phi0?mu0?v0,80 @ phi*?mu*?v*,8*+ ? 5c <-= d&e!+,d&t+ > alpha!?5g,&:!@5g+?u! & beta!@ sigmamuv+ ? e! @ alphastar </= d&e0+,d&t+ > alpha0?5e,&:0@5e+?u0 & beta0@ sigmamuv+ ?e0 @ alphastar <(= d&e*+,d&t+ > alpha*?5g,&:*@5g+?u* & beta*@ sigmamuv+ ?e* @ alphastar <6= d&5t+,d&t+ > d5t

7xplicit equations as entered by the user <!= beta! > .( <0= gam* > .6 <*= mu!max > .)) <)= :o0 > .4! <-= beta0 > .( </= beta* > .( <(= mu0max > .!9 <6= mu*max > .*/ <9= alpha* > .* <!4= :o* > 0.0 <!!= alpha! > .* <!0= alpha0 > .* <!*= e*max > alpha*,& mu*max@ beta*+ <!)= e!max > alpha!,& mu!max@ beta!+

<!-= e0max > alpha0,& mu0max@ beta0+ <!/= 7* > e*,e*max <!(= 7! > e!,e!max <!6= 70 > e0,e0max <!9= :! > .4<04= :0 > .4! <0!= :* > .44! <00= mu! > mu!max?7!?5g,&:!@5g+ <0*= mu0 > mu0max?70?5e,&:0@5e+ ? 5o,&:o0 @ 5o+ <0)= mu* > mu*max?7*?5e,&:*@5g+ ? 5o,&:o* @ 5o+ <0-= mumax > if &mu!Emu0+ then &if &mu!Emu*+ then &mu!+ else &mu*++ else &if &mu0Emu*+ then &mu0+ else&mu*++ <0/= u! > mu!,&mu!@mu0@mu*+ <0(= u0 > mu0,&mu!@mu0@mu*+ <06= 8! > .!/ <09= 80 > .(<*4= 8* > ./4 <*!= u* > mu*,&mu!@mu0@mu*+ <*0= alphastar > .! <**= 5g4 > !4 <*)= 5s04 > 04 <*-= ; > 4 <*/= v* > mu*,mumax <*(= v! > mu!, mumax

<*6= v0 > mu0,mumax <*9= sigmamuv > mu!?v!@mu0?v0@mu*?v* <)4= d5c > &sigmamuv - ;+?5c <)!= gam! > !4 <)0= gam0 > !4 <)*= d5t > gam*?mu*?v* &gam!?mu!?v!@gam0?mu0?v0+?5t sigmamuv?5t <))= phi! > .)6 <)-= phi0 > 0 <)/= phi* > ! <)(= phi) > .9<)6= "la > !444 <)9= 5ostar > .! 5lic" here for the Aolymath 5ode

igure ). 5ybernetic model simulations and experimental data from von Meyenburg &!9/9+ for cell mass, glucose, ethanol concentrations in aerobic batch culture of Saccha"omycesce"evisiae.

;uring the first growth phase, the yeast cells clearly prefer the faster fermentative metabolism and ignore or repress the oxidative metabolism. This choice of the fermentative pathway can be concluded from &!+ the growth rate &the slope of a semi-long plot of cell mass+ during the first growth phase or more easily &0+ the accumulation of the fermentation product, ethanol. 2fter glucose is completely fermented, the presence of oxygen enables further growth of yeast cells in a second or diauxic growth phase using ethanol oxidative pathway.

The yeast cybernetic model equations for continuous or chemostat cultures can be simulated to predict the C5rabtree effectC of preferential utilization of glucose oxidative pathway during the low dilution rates, followed by switch to the fermentative pathway at higher dilution rates. The oxidative consumption of glucose, which is not utilized in the batch aerobic cultures, is the preferred pathway for glucose consumption at the low dilution rates, as seen from the high cell mass yields in igure - and absence of any ethanol production. 2t higher dilution rates, the utilization of glucose fermentative pathway is seen both in low cell mass yield in igure - as well as the production of ethanol &data not shown+. #e can use the earlier program, and set different values for ; to get various concentration plots of the cell and the substrates. Diven are the tables of values for ; > .0, .*, .) P+,-M./) /.B,E0 Calculated !alues of the DE1 !ariables 3D @ . >6

3ariable initial value minimal value maximal value final value t 4 4 04 4.49/*)/9 04 *(.*0!(

5c 4.! *(.*0!( 5g ) -004.(4*)) 5e 04 !!4.)6(6( 5o 4.4.499)06* e! 6.*7-44.*4-90( e0 !.47-4/ 4.!4-60*9 e* !.47-4) 4.46!!)!) 5t 4 !.*67-4/ beta! gam* 4.6 mu!max 4.)) :o0 4.4! beta0 beta* mu0max 4.!9 mu*max 4.*/ 4.( 4.6

-004.(4*))

/.!004*4*

/.*-0*040

!!4.)6(6(

4.499)06*

4.-

6.*7-4-

4.*4(6-)(

!.47-4/

4.!466*(

!.47-4)

4.!40!9*-

4.4*/4-90

4.( 4.6

4.( 4.6

4.(

4.))

4.))

4.))

4.4!

4.4!

4.4!

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.(

4.*/

4.*/

4.*/

alpha* :o* alpha! alpha0

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.* 4.06*4!69

4.* 0.0 4.* 4.*

e*max 4.06*4!69 4.06*4!69 4.06*4!69 e!max 4.0/*!-(9 4.0/*!-(9 4.0/*!-(9 e0max 4.**(4(6( 4.**(4(6( 4.**(4(6(

4.0/*!-(9

4.**(4(6(

7* *.-**7-4) *.-**7-4) 4.*/!)9(( 4.06//996 7! *.!-)7-4) *.!-)7-4) !.!/96)6! !.!/0-00( 70 0.9/(7-4/ 0.9/(7-4/ 4.*006(-) 4.*!*9))* :! 4.4:0 4.4! :* 4.44! 4.44.44.4-

4.4!

4.4!

4.4!

4.44!

4.44!

4.44!

mu! !.*(!7-4) !.*(!7-4) 4.-(60)/) 4.-!!/0-9 mu0 -.-0*7-4( -.-0*7-4( 4.4--/6*( 4.4-)!9*mu* !.!((7-4) -4.!/!/)0/ 4.!!-9!-0 -4.4400*)0 mumax 4.-(60)/) !.*(!7-4) 4.-!!/0-9 !.*(!7-4)

u! 4.-*/(*/ 4.94(64-6

4.-*/(*/

!.0*!(*/!

u0 4.440!/09 4.440!/09 4.!!0-60/ 4.49/!-68! 4.!/ 80 4.(8* 4.!/ 4.!/ 4.!/

4.(-

4.(-

4.(-

4./

4./

4./

4./

u* 4.)/!!4!! -4.*))*!6( 4.)/!!4!! -4.44*9/)* alphastar 4.! 5g4 5s04 ; 4.! 4.! 4.!

!4 04 4.0

!4 04 4.0

!4 04 4.0

!4 04 4.0

v* 4.6-946*( -4.0(9-*9* 4.6-946*( -4.44)*//9 v! ! ! ! !

v0 4.44)4096 4.44)4096 4.!*064( 4.!4-90)! sigmamuv 4./060/0( 0.*607-4) 4.-!(*(/ 0.*607-4)

d5c -4.4!99(/0 -4.4!99(/0 !!.6)-4!* !!.6)-4!* gam! !4 gam0 !4 !4 !4 !4

!4

!4

!4

d5t 6.4907-4-4.!9(6*// 4.4!(940- -).9)/7-46 phi! 4.)6 phi0 4.)6 4.)6 4.)6

phi* phi) 4.9"la !444 5ostar

! 4.9-

! 4.9-

! 4.9-

!444

!444

!444

4.!

4.!

4.!

4.!

Calculated !alues of the DE1 !ariables 3D @ . &6

3ariable initial value minimal value maximal value final value t 4 4 04 4.49!)9!6 04 ).6)60)4/

5c 4.! ).6)60)4/ 5g ) -!9./9*6*/ 5e 04 !).!(/)(! 5o 4.4.49990/( e! 6.*7-44.*4//*99 e0 !.47-4/ 4.!4-*)6e* !.47-4) 4.464)!(5t 4 0.94(7-4/ beta! gam* 4.6 4.( 4.6

-!9./9*6*/

(./!/464!

0./*/(9)!

04

4.496*6/

4.-

6.*7-4-

4.*4(0-/*

!.47-4/

4.!!4*!)*

!.47-4)

4.!0)64!

4.4-!9(--

4.( 4.6

4.( 4.6

4.(

mu!max 4.)) :o0 4.4! beta0 beta* mu0max 4.!9 mu*max 4.*/ alpha* :o* alpha! alpha0

4.))

4.))

4.))

4.4!

4.4!

4.4!

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.(

4.*/

4.*/

4.*/

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

e*max 4.06*4!69 4.06*4!69 4.06*4!69 e!max 4.0/*!-(9 4.0/*!-(9 4.0/*!-(9 e0max 4.**(4(6( 4.**(4(6( 4.**(4(6( 7* *.-**7-4) 4.06)!)!(

4.06*4!69

4.0/*!-(9

4.**(4(6(

*.-**7-4)

4.))4)49

7! *.!-)7-4) *.!-)7-4) !.!/(-(** !.!/-0*!/ 70 0.9/(7-4/ 0.9/(7-4/ 4.*0(0/4( 4.*!0-**9 :! 4.4:0 4.4! 4.44.44.4-

4.4!

4.4!

4.4!

:* 4.44!

4.44!

4.44!

4.44!

mu! !.*(!7-4) !.*(!7-4) 4.-!(/-4/ 4.-!)44/9 mu0 -.-0*7-4( -.-0*7-4( 4.4-/*9(9 4.4-*9)!mu* !.!((7-4) -4.4(!60-6 4.-//09(( -4.44*!99) mumax 4.-//09(( !.*(!7-4) 4.-!)44/9 !.*(!7-4)

u! 4.-*/(*/ 4.0/0/60!.4*96!00 4.9!4!-!! u0 4.440!/09 4.440!/09 4.!46(9)* 4.49--!)! 8! 4.!/ 80 4.(8* 4.!/ 4.!/ 4.!/

4.(-

4.(-

4.(-

4./

4./

4./

4./

u* 4.)/!!4!! -4.!)6/4/4./(0*!4) -4.44-//-! alphastar 4.! 5g4 5s04 ; 4.! 4.! 4.!

!4 04 4.*

!4 04 4.*

!4 04 4.* !

!4 04 4.*

v* 4.6-946*( -4.44/00)) v! ! !

-4.!)09!/(

4.*94(!/0

v0 4.44)4096 4.44)4096 4.!-/(09( 4.!4)9)*!

sigmamuv 4./-64)0(

0.*607-4) 4.-!9/6(/

0.*607-4)

d5c -4.4099(/0 -4.4099(/0 !.4/-496* !.4/-496* gam! !4 gam0 !4 !4 !4 !4

!4

!4

!4

d5t 6.4907-4-4.009*-*6 4.*(!6/-6 -/.6*)7-4( phi! 4.)6 phi0 phi* phi) 4.9"la !444 5ostar 4.)6 4.)6 4.)6

0 ! 4.9-

0 ! 4.9-

0 ! 4.9-

0 !

!444

!444

!444

4.!

4.!

4.!

4.!

Calculated !alues of the DE1 !ariables 3D @ . 56

3ariable initial value minimal value maximal value final value t 4 4 04 4.46*9)*) 04 4./)0*4/*

5c 4.! 4./)0*4/* 5g ) /.409*-09 5e 04 !.94!9*0-

6.()9/()*

4.9*06(-6

04

5o 4.4.49999 e! 6.*7-44.*4)!)( e0 !.47-4/ 4.!4/))-! e* !.47-4) 4.46*!/6/ 5t 4 -.-6(7-4( beta! gam* 4.6 mu!max 4.)) :o0 4.4! beta0 beta* mu0max 4.!9 mu*max 4.*/ alpha* :o* alpha! alpha0 4.( 4.6

4.49999

4.-

6.*7-4-

4.*4)(!6!

!.47-4/

4.!4/(49!

!.47-4)

4.46)9*-

/.!/!7-4-

4.( 4.6

4.( 4.6

4.(

4.))

4.))

4.))

4.4!

4.4!

4.4!

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.(

4.*/

4.*/

4.*/

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.* 4.06*4!69

4.* 0.0 4.* 4.*

e*max 4.06*4!69 4.06*4!69 4.06*4!69 e!max 4.0/*!-(9 4.0/*!-(9 4.0/*!-(9

4.0/*!-(9

e0max 4.**(4(6( 4.**(4(6( 4.**(4(6(

4.**(4(6(

7* *.-**7-4) *.-**7-4) 4.*44!4*0 4.09*6/0* 7! *.!-)7-4) *.!-)7-4) !.!-(9066 !.!--(-6/ 70 0.9/(7-4/ 0.9/(7-4/ 4.*!/-(4) 4.*!-(6(0 :! 4.4:0 4.4! :* 4.44! 4.44.44.4-

4.4!

4.4!

4.4!

4.44!

4.44!

4.44!

mu! !.*(!7-4) !.*(!7-4) 4.-4/)946 4.-4)*-!* mu0 -.-0*7-4( -.-0*7-4( 4.4-)--!! 4.4-)0-9* mu* !.!((7-4) !.!((7-4) 4.44666/( 4.44!)-4mumax 4.-4/)946 !.*(!7-4) 4.-4)*-!* !.*(!7-4)

u! 4.-*/(*/ 4.-*/(*/ 4.94*0//* 4.944-09 u0 4.440!/09 4.440!/09 4.49(9(0( 4.49/66! 8! 4.!/ 80 4.(8* 4.!/ 4.!/ 4.!/

4.(-

4.(-

4.(-

4./

4./

4./ 9.4!*7-4)

4./

u* 4.)/!!4!! 4.)/!!4!! 4.440-9

alphastar 4.! 5g4 5s04 ;

4.!

4.!

4.!

!4 04 4.)

!4 04 4.)

!4 04 4.)

!4 04 4.)

v* 4.6-946*( 9.9(67-4) 4.6-946*( 4.4406(/ v! ! ! ! !

v0 4.44)4096 4.44)4096 4.!!00009 4.!4(-60* sigmamuv 4.-!0!9(! 0.*607-4) 4.-!4!906 0.*607-4)

d5c -4.4*99(/0 -4.4*99(/0 4.4(4(((4.4(4(((gam! !4 gam0 !4 !4 !4 !4

!4

!4

!4

d5t 6.4907-44) 0.4007-4( phi! 4.)6 phi0 phi* phi) 4.9"la !444 5ostar 4.)6

--.0607-4-

!.!!!7-

4.)6

4.)6

0 ! 4.9-

0 ! 4.9-

0 ! 4.9-

0 !

!444

!444

!444

4.!

4.!

4.!

4.!

Figure ,. (hemostat or continuous culture steady state data on cell mass yield on glucose

The transition from the oxidative to fermentative pathways occurs either gradually or abruptly, depending on other culture conditions, such as oxygen supply rates, controlled mainly by impeller agitation rates, which are different in these experimental studies.

2.7.4.4 Metabolic +scillations : C"bernetic Model

2t the intermediate dilution rates, as the cells are changing from oxidative to fermentative metabolism, yeast cells can exhibit s#ontaneous metabolic oscillations over a range of operating conditions, such as the agitation rate or oxygen mass transfer rate. 1everal experimentalists have documented these oscillations in aerobic continuous cultures of yeast on glucose. The figure from Aorro et al. &!966+ shows the sustained oscillations in all

the metabolite concentrations measured in the continuous bioreactor. The top panel shows the online measurement trace of dissolved oxygen concentration in the culture medium, which is most readily obtained. The subsequent panels show off-line measurements of ethanol, glucose, intracellular storage carbohydrates &trehalose and glycogen+, medium pG, and cell number concentration &L,ml+. These spontaneous oscillations change in shape, period and amplitude as the bioreactor operating conditions of dilution rate and agitation rate are varied within their oscillatory ranges. 2s these operating conditions are varied outside their oscillatory range, the oscillations die down to either oxidative or fermentative consumption of glucose. The metabolic oscillations can also be predicted by the yeast cybernetic model equations given above &Kones and :ompala, !999+ but requires the use of a stiff M;7 solver. The simulations were conducted on the Ner"eley Madonna software at the operating conditions of dilution rate of 4.!/ hr-! and oxygen mass transfer rate -.a &strongly affected by the agitation rate+ of *44 hr-!. The parameter values for the different model constants are listed in the Table '.(.).!. 2t different operating conditions, the cybernetic model predicts also the experimentally observed trends in the shape and period of oscillations as well as the damping of the oscillations to either the fermentative or the oxidative consumption outside the range of oscillatory conditions.

P+,-M./) /.B,E0

3ariable initial value minimal value maximal value final value t 4 4 4.! 04 04

5c 4.! !64!.0)*! 5g 4.! -!.!!-7@4) 5e 4.! -*06.--!/ 5o !.0 4.4*---/e! 6.*7-44.*4(-9*! e0 !.47-4) 4.!4*)-)6 e* !.47-4) 4.46!/*09 5t 4.! 0.4007-4( beta! gam* 4.6 mu!max 4.)) :o0 4.4! beta0 beta* mu0max 4.!9 4.( 4.6

!64!.0)*!

-!.!!-7@4)

4.!

4.!

-*06.--!/

4.4*---/-

!.0

6.*7-4-

4.*4((!*

!.47-4)

4.!-9!)00

!.47-4)

4.!-/!0*9

0.4007-4(

4.!

4.( 4.6

4.( 4.6

4.(

4.))

4.))

4.))

4.4!

4.4!

4.4!

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.( 4.!9

4.( 4.(

mu*max 4.*/ alpha* :o* alpha! alpha0

4.*/

4.*/

4.*/

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.*

4.* 0.0 4.* 4.* 4.06*4!69

4.* 0.0 4.* 4.*

e*max 4.06*4!69 4.06*4!69 4.06*4!69 e!max 4.0/*!-(9 4.0/*!-(9 4.0/*!-(9 e0max 4.**(4(6( 4.**(4(6( 4.**(4(6(

4.0/*!-(9

4.**(4(6(

7* *.-**7-4) *.-**7-4) 4.--440** 4.066)*/* 7! *.!-)7-4) *.!-)7-4) !.!/9*4(( !.!/66-*9 70 0.9/(7-4) 0.9/(7-4) 4.)(!))() 4.*4/9!/ :! 4.4:0 4.4! :* 4.44! 4.44.44.4-

4.4!

4.4!

4.4!

4.44!

4.44!

4.44!

mu! 9.0-07-4-4.4940/(4.-*69460 4.-!)096 mu0 -.4607-4-.4607-44.4(-996) 4.4)--!*/ mu* ).))-7-4-4.49/*!( 4.!4*/49- -(.69*7-4) mumax 4.-*69460 9.0-07-44.-!)096 9.0-07-4-

u! 4.)90/()/ 4.!(4/)(( 4.9!999-/ 4.9!999-/ u0 4.0(4/0!( -4.-64)6!) 4.*(6/!(9 4.46!)!/) 8! 4.!/ 80 4.(8* 4.!/ 4.!/ 4.!/

4.(-

4.(-

4.(-

4./

4./

4./

4./

u* 4.0*/(4*( -4.4!9!909 4.6!-6/00 -4.44!)!0 alphastar 4.! 5g4 5s04 ; 4 4.! 4.! 4.!

!4 04 4

!4 04 4

!4 04 4 !

!4 04

v* 4.)64))/) -4.44!-*)6 v! ! !

-!.)4-)90)

-!.*!(0!-6

v0 4.-)90949 4.466)9/sigmamuv 4.-)-(06/ !.)!67-4) 4.-!6*0(

4.466)9/-

!.)!67-4)

d5c !.)!67-4!.)!67-49**./*09) 9**./*09) gam! !4 gam0 !4 !4 !4 !4

!4

!4

!4

d5t -!.!(-7-4) -4.46(()-! 4.4/0!96 -!.6*67-4(

phi! 4.)6 phi0 phi* phi) 4.9"la *44 5ostar

4.)6

4.)6

4.)6

0 ! 4.9-

0 ! 4.9-

0 ! 4.9-

0 !

*44

*44

*44

4.!

4.!

4.!

4.!

Figure !. (ybernetic model simulations of metabolic oscillations in yeast continuous cultures. 2he shape and period of the oscillations in all the metabolite concentrations agree -ualitati)ely with the experimental data. /s the bioreactor operating conditions of D and kLa are changed, the shape and period of oscillations change as well in both experimental data and model simulations.

/able 2.7.4.8 ?omenclature and Parameter Balues used in Model E9uations and Simulations

Aarameter

.nits

;efinition

3alue

mmax,# mmax,2 mmax,3 9# 92 93

:M0 :M* 8! 80 8* ai a?i bi g! g0 g* f! f0 f* f) 5D,4, 514, 55 51, 51!, 510 5D, 57,, 5M

hr

4#

hr4# hr4# g:; g:; g:;

g,F g,F g,g g,g g,g hr-! hr-! hr-! g,g g,g g,g g,g g,g g,g g,g g,F g,F g,F g,F

3ax. %pecific 5rowth ate for 5lucose 1."" Fermentation 3aximum %pecific 5rowth ate for Ethanol 1.#7 6xidation 3aximum %pecific 5rowth ate for 5lucose 1.38 6xidation 3onod %aturation (onstant for 5lucose 1.1, Fermentation 3onod %aturation (onstant for Ethanol 6xidation 1.1# 3onod %aturation (onstant for 5lucose 6xidation 1.11# 6xygen %aturation (onstant for Ethanol 6xidation 4.4! 6xygen %aturation (onstant for 5lucose 6xidation 0.0 4.!/ <ield (oefficient for 5lucose Fermentation 4.(<ield (oefficient for Ethanol 6xidation 4./4 <ield (oefficient for 5lucose 6xidation 4.* Enzyme %ynthesis ate (onstant 4.! (onstituti)e Enzyme %ynthesis ate (onstant 4.( Enzyme Degradation ate (onstant !4 ate (onstant for (arbohydrate Degradation !4 ate (onstant for (arbohydrate Degradation 4.6 ate (onstant for (arbohydrate %ynthesis %toichiometric (oefficient for Ethanol +roduction 4.)6 %toichiometric (oefficient for Ethanol 6xidation 0 %toichiometric (oefficient for 5lucose 6xidation ! 4.9%toichiometric (oefficient for (arbohydrate +roduction =nlet or Feed 5lucose and %ubstrate (oncentrations!4, var.
5ellmass 5oncentration 1ubstrate 5oncentration Dlucose, 7thanol, and ;issolved Mxygen variable variable variable

concentrations 5T e%, emax,i 7i mi ; u v "Fa g,g g,g hr hr hr

! ! !

%ntracellular 1torage 5arbohydrate &Trehalose+ 5onc. %ntracellular 7nzyme 5oncentration and its maximum 'elative 2mount of %ntracellular 7nzyme Drowth 'ate on iHth 1ubstrate or Aathway ;ilution 'ate 5ybernetic 3ariables 5ontrolling 7nzyme 1ynthesis Mass transfer coefficient for ;issolved Mxygen

variable variable variable variable variable variable

5ybernetic 3ariables 5ontrolling 7nzyme 2ctivity variable variable

2eferences0

!. :ompala, ;.1., ;. 'am"rishna, I.N. Kansen and D.T. Tsao, C%nvestigation of bacterial growth on mixed substrates. 7xperimental evaluation of cybernetic models,C Biotechnolog" and Bioengineering >$:!4))-!4-- &!96/+.

0. Kones, :.;. and ;.1. :ompala, C5ybernetic model of the growth dynamics of Saccha"omyces ce"evisiae in batch and continuous cultures,C C. Biotechnolog" 780 !4--!*! &!999+.

)omework Problems Based on this Sample Section 0

!. 1how that the enzyme modification of the Monod growth "inetics is capable of simulating the presence or absence of the initial lag phase by varying the initial level of the intracellular enzyme content.

0. 1how that the enzyme modification of Monod growth "inetics does not affect the chemostat profiles of cell mass and substrate over the range of dilution rates as well as the washout and optimal dilution rates.

*. 7xamine whether the order of substrate preference is affected by the choice of initial levels of the two "ey combinations &test the four combinations: high-high, high-low, low-high, and low-low where high represents 99O of emax,i and low represents !O of emax,i+.

). Eshe"ichia coli grows on a mixture of three sugars: glucose, xylose and lactose. The Monod growth rate parameters for glucose are: mmax,D > !.0 hr-!, :D > 4.4! g,F, 85,D > 4.-- g,gP for xylose: mmax,Q > !.4- hr-!, :Q > 4.4- g,F, 85,Q > 4.-0 g,gP and for lactose: mmax,F > 4.6 hr-!, :F > 4.! g,F, 85,F > 4.)- g,g. Gow will these sugars be consumed in a batch bioreactor with the initial sugar concentrations of - g,F glucose, !4 g,F xylose, and 0- g,F of lactose. 7xtend the cybernetic model framewor" to address three substrates.

-. E. coli is to be cultured in a chemostat on a mixture of glucose, xylose and lactose with the feed concentrations of - g,F glucose, !4 g,F xylose, and 0- g,F of lactose. The Monod growth rate parameters are same as the ones in the problem above. #hat is the optimal dilution rate that will maximize the cell mass production rate &;?5c+J

/. /ymomonas mobilis has been engineered to ferment pentoses li"e xylose in addition to the common hexose, glucose. The Monod parameters for this metabolically engineered microorganism during growth on glucose are mmax > 4.)4 hr-!, :s > 4.! g,l, 8x,s > 4.!! gdw , g1, and 8p,s > 4.)6 g ethanol , g glucose. The same parameters for growth on xylose are mmax > 4.*4 hr-!, :s > 4.- g,l, 8x,s > 4.!4 gdw,g1, and 8p,s >

4.)- g ethanol,g xylose. These cells are grown in a chemostat fermentor with the feed containing -4 g,l glucose and -4 g,l xylose. .sing the Monod chemostat equations, calculate the maximum ethanol production rate possible from these cells growing in a chemostat. 2ssume that the fermentative growth on xylose gets repressed or shut off completely when the glucose concentration exceeds 4.0 g,l and that both sugars are fermented at lower dilution rates or glucose concentration R 4.0 g,l. Ma"e any other assumptions as needed.

(. 1olve the above problem using the cybernetic model equations, using a > 4.444! and b > 4.4- for both "ey enzymes and without using the assumptions stated in the last two sentences of above problem.

6. 5ontinuing with the theme of above problem, a second chemostat &of the same size+ is added in series or downstream from the first chemostat, operated at high dilution rate, to ensure that all of xylose is consumed in continuous culture,. #hat dilution rate should be used to maximize the ethanol production rate from the metabolically engineering /ymomonas mobilisJ .se the growth parameters from the earlier problem statement. .se the cybernetic model equations or state your additional assumptions.

0. Saccha"omyces ce"evisiae grows in a typical diauxic growth phenomenon on a mixture of glucose and galactose. The Monod growth rate parameters for the fermentative growth on galactose are assumed to be mmax > 4.)4 hr-!, :s > 4.! g,F, 8x,s > 4.!- g cell mass , g galactose, and 8p,s > 4.)( g ethanol , g galactose. The growth rate parameters for oxidative growth on galactose are assumed to be mmax > 4.** hr-!, :s > 4.44! g,F, :M* > 0.- g,F and 8x,s > 4.-6 g cell mass , g galactose. 1imilar parameters for growth on glucose and ethanol are given in Table '.(.).!. ;etermine how the cell mass, glucose, galactose and ethanol profiles will be in batch culture on a mixture of !4 g,F glucose and 04 g,F galactose, if the inoculum is

ta"en from continuous cultures on a mixture of glucose and galactose at a dilution rate.

11. 5ontinuing with the theme of the above problem, S. ce"evisiae is grown in a chemostat on a mixture of glucose and galactose, with the feed concentration of each being -4 g,F. %t is desired to maximize the cell mass production rate &;?5c+. #hat should be the dilution rate used in the single chemostat, assuming a -.a of !444 hr-!J #atch out for the possibility of spontaneous metabolic oscillations.

11. 5ontinuing with the theme of the above problem further, S. ce"evisiae is grown in a chemostat on a mixture of glucose and galactose, with the feed concentration of each being -4 g,F, to maximize the ethanol production rate &;?57+. #hat should be the dilution rate used in the single chemostat, assuming a -.a of !44 hr-!J

!0. 1imulate the metabolic oscillations of yeast, using the cybernetic model parameters in Kones and :ompala &!999+ &given in the Table '.(.).!+ to &a+ Alot how the period of oscillations changes with dilution rate , and &b+ Alot how the period changes with the mass transfer coefficient, -.a. .se a sti!! equation solver in your simulations to obtain the oscillations.

!*. %t has been found experimentally that the spontaneous metabolic oscillations in continuous cultures of yeast S. ce"evisiae can be avoided by adding a small amount of ethanol to the feed stream, along with glucose feed concentration of *4 g,F. %nvestigate whether the cybernetic model equations given in section '.(.).) can predict the elimination of oscillations with the inclusion of ethanol in the feed stream. #hat is the smallest ethanol concentration that will eliminate the oscillationsJ

!). ;etermine the parameter sensitivity of the cybernetic model simulations of spontaneous metabolic oscillations in continuous cultures of yeast for the following assumed parameter values: a, a?, b, mmax,(, :*, :M* and 8*. irst obtain the oscillations through numerical simulations of the cybernetic model for any combination of bioreactor operating parameters, , and -.a. Iext vary each of the assumed parameters to determine if and how the shape and period of oscillations change from the base case.

,aborator" E;ercise0 -east *ermentation

86 =ntroduction

%n this laboratory exercise, we will study the growth characteristics of the yeast Saccha"omyces ce"evisiae in batch cultures. 2 labscale &- liter+ fermentor will be used to study batch growth "inetics of the yeast growing on glucose as the single carbon substrate provided in the presence of oxygen. 2 second fermentor will also contain glucose as the sole carbon substrate for the yeast to utilize in the absence of oxygen. 2 third lab-scale fermentor will be used to observe the growth behavior when the cells are presented with a mixture of two carbohydrates, glucose and glycerol in the presence of oxygen. 8ou will ta"e samples from the fermentors, measure the cell mass concentration &through optical density+ and determine the concentrations of glucose and ethanol with spectrophotometric assay "its. #ith the accumulated results from the all the students over eighteen hours for the three different fermentors, you will be able to analyze the "inetics of cell growth, and the different patterns of multiple substrate utilization in batch cultures.

-east Metabolism

Saccha"omyces ce"evisiae uses the following three ma$or pathways for growth on glucose:

!+ The fermentation of glucose, which occurs primarily when the glucose concentration is high or when oxygen is not available. The cells attain a maximum specific growth rate of about 4.)- hr-! with a low biomass yield of 4.!- g dry mass per gram glucose consumed and a high respiratory quotient &the ratio of 5M0 production rate to the M0 consumption rate+ and a low energy yield of only about 0 2TA per mole of glucose metabolized. The stoichiometry of this reaction is

@ 05M0 @ e

5/G!0M/ --------------E

050G-MG

where e represents chemical energy utilized in the growth processes.

0+ The oxidation of glucose, which predominates at glucose concentrations below -4 mg,l in aerobic cultures. The cells attain a maximum specific growth rate of only about 4.0- hr-! with a biomass yield of about 4.- g dry mass per gram glucose consumed, a respiratory quotient of about !, and a high energy yield of !/-06 2TA per mole of glucose metabolized. The stoichiometry of this reaction is:

5/G!0M/ @ /M0 /5M0 @ /G0M @ e

--------------E

*+The oxidation of ethanol, which predominates when fermentative substrates are not available or in very limited supply. The cells attain a maximum specific growth rate of about 4.0 hr-! with a high biomass yield of about 4./-4.( g dry mass per gram ethanol consumed, a low respiratory quotient of about 4.(, and an energy yield of about /-!! 2TA per mole of ethanol metabolized. The stoichiometry of this reaction is:

50G-MG @ *M0 05M0 @ *G0M @ e

--------------E

.tilization of glycerol &the second carbon substrate provided in the third fermentor+ by Saccha"omyces ce"evisiae is repressed by glucose. 2fter the depletion of the faster growth-supporting substrate, glucose, the enzymes necessary for the assimilation of glycerol is induced, and an exponential growth phase on glycerol is expected to follow a diauxic lag phase.

E;perimental Conditions

The temperature of the water bath surrounding the fermentors will be controlled at *4S5, and the impellers inside each fermentor will be operated at )44 rpm. Two of the three fermentors will have glucose as the only initial carbon source at a concentration of !4 g,l. The other batch fermentor will have two initial carbon sources - glucose at * g,l and glycerol at ( g,l. 2ir will be sparged into the first and the third fermentors at constant rate of !4 l,min. The dissolved oxygen concentration in all the three fermentors can be computer controlled to maintain a desired level by enriching the oxygen concentration in the sparged gas.

=ntroduction to ,ab Procedures

8ou will monitor the growth characteristics of the yeast in the batch fermentors in three ways by measuring the concentration of cells, and preparing cell-free samples at hourly intervals from each bioreactor for assaying the concentrations of glucose and ethanol. 8east cell concentration can be determined indirectly by measuring the optical density &absorbance+ of a culture sample. 8ou will ta"e a sample of the culture medium from the fermentor and read its absorbance using a spectrophotometer. .p to a certain cell density, the concentration of yeast cells &gdw,l+ in the sample is proportional to the absorbance reading on the spectrophotometer. The calibration curve correlating cell concentration with absorbance deviates from a linear correlation at high cell densities. Necause of this, itHs a good idea to dilute any of your high M; samples &that may be on the non-linear portion of the curve+ by a "nown dilution factor to confirm that the measured M; values fall on the linear portion.

The concentration of glucose, &glycerol+ and ethanol in the cell-free culture samples will be analyzed by high performance liquid chromatography or spectrophotometric assay. To remove the cells from a * ml sample, the cell suspension will be centrifuged and the supernatant will be filtered through a microfiltration syringe, and assayed for glucose and ethanol by the spectrophotometric assay.

2) Experimental Procedures

.6 Determining Cell Concentration

!+ Teroing the spectrophotometer. 1et the wavelength to /*4 nm. .sing the "nob on the left, set the reading to 4O transmission when the chamber is emptyP using the "nob on the right, set the reading to !44O transmission when the

chamber contains a test tube with about ) ml of pure medium.

0+ ;etermining cell concentration. irst flush out the sample tube for your groupHs fermentor by ta"ing an 6-!4 ml sample, which you will then discard. Ta"e another 6-!4 ml sample from your groupHs fermentor and gently mix. Ta"e about ) ml from your sample tube and transfer it to a glass test tube. 5lean the outside of the test tube with ethanol, insert it into the spectrophotometer, and record the absorbance reading. %f the absorbance reading is greater than 4.0-, a typical limit of linear correlation between the absorbance and cell mass concentration, dilute the sample with a "nown amount of pure medium, and measure the absorbance again to chec" if the absorbance reading is on the linear portion of the calibration curve. 'ecord the time you ta"e the sample along with the absorbance reading in the linear range as well as the dilution factor.

B6 Spectrophotometric .ssa"s for Glucose and Ethanol

To remove the cells from a * ml sample, the culture will be centrifuged and the supernatant will be filtered through a microfiltration syringe. The lab T2 will provide more details on the assay procedures during the experiment.

&6 2eport : Due *ebruar" 8'# 8''7

2+ ;raw a graph of: a+ Fogarithm of cell concentration

vs. time

b+ Dlucose concentration vs. time c+ 7thanol concentration vs. time

for each of the three batch fermentors.

N+ ;etermine the specific growth rate and the yield coefficient &gram dry weight of cells produced per gram of carbon source consumed+ for each growth phase in the three fermentors. &The calibration between the absorbance reading and the dry cell mass concentration of the yeast cells will be performed at the end of the batch cultures and provided in the following class period+.

5+ %nterpret these two graphs in light of the bac"ground information on yeast metabolic pathways and the CcyberneticC principle that cells choose to grow at the fastest possible rate. 1pecifically, discuss why the cell mass, glucose, &glycerol+ and ethanol concentration profiles loo" as they do for each batch fermentor.

ogler U Durmen V 0446 .niversity of Michigan