Resea ch a icle
Tailo ing non-axenic lac ic acid e men a ion om cheese whey pe mea e
a ge ing a lexible lac ic acid pla o m
B ech Delmoi i´
e
a,b
, My sini Saka ika
a,b
, Ko neel Rabaey
b,c
, Heleen De We e
b,d
,
Albe e Reguei a
a,b,e,*
a
Cen e o Mic obial Ecology and Technology (CMET), Facul y o Bioscience Enginee ing Ghen Uni e si y, Coupu e Links 653, 9000, Ghen , Belgium
b
Cen e o Ad anced P ocess Technology o U ban Resou ce Reco e y (CAPTURE), F ieda Saeyss aa 1, 9052, Ghen , Belgium
c
Depa men o Bio echnology, Facul y o Bioscience Enginee ing, Ghen Uni e si y, Coupu e Links 653, 9000, Ghen , Belgium
d
Flemish Ins i u e o Technological Resea ch, (VITO), Boe e ang 200, 2400, Mol, Belgium
e
CRETUS, Depa men o Chemical Enginee ing, Uni e sidade de San iago de Compos ela, 15075 San iago de Compos ela, Spain
ABSTRACT
Lac ic acid (LA) is an impo an biobased pla o m chemical, wi h po en ial applica ions in syn he ising a wide ange o chemical p oduc s o se ing as eeds ock o
a ious biop ocesses. Indus ial LA p oduc ion ia pu e cul u e e men a ion is cha ac e ized by high ope a ional cos s and u ilizes ood-g ade suga s, he eby
educing he easibili y o LA applica ions. In his con ex , ou esea ch ocussed on alo ising he la ges dai y side s eam, cheese whey pe mea e, h ough he use o
mixed mic obial communi ies. We e alua ed he e ec o di e en ope a ional pa ame e s ( empe a u e, pH and hyd aulic e en ion ime) in non-axenic e men-
a ions on p oduc i i y, yield, concen a ion, op ical pu i y, and communi y. Ou indings e ealed ha ope a ing a mildly he mophilic condi ions (45 ◦C) esul ed
in highly selec i e LA p oduc ion, and signi ican ly augmen ed he LA yield, and p oduc i i y, compa ed o highe empe a u es (50–55 ◦C). In addi ion, ope a ing a
ci cumneu al pH condi ions (6.0–6.5) led o signi ican ly inc eased he LA e men a ion pe o mance compa ed o he con en ional acid pH condi ions (≤5.5). This
led o an unp eceden ed LA p oduc i i y o 27.4 g/L/h wi h a LA yield o 70.0% which is 2.5 imes highe compa ed o p e ious epo ed maximum. Addi ionally,
a ying pH le els in luenced he op ical pu i y o LA: we achie ed an op ical L-LA pu i y o 98.3% a pH 6.0–6.5, and an op ical D-LA pu i y o 91.3% a a pH o 5.5. A
sho hyd aulic e en ion ime o less han 12 h was c ucial o selec i e LA p oduc ion. This p ocess also yielded a mic obial biomass composed o 90.3–98.6%
Lac obacillus delb ueckii, which could be po en ially alo ised as p obio ic o p o ein ing edien in ood o eed p oduc s. Ou wo k shows ha by ca e ul selec ion o
ope a ional condi ions, he o e all pe o mance can be signi ican ly inc eased compa ed o he s a e-o - he-a . These esul s highligh he po en ial o non-s e ile LA
e men a ion and show ha ca e ul selec ion o simple eac o ope a ion pa ame e s can maximize p ocess pe o mance. A p elimina y assessmen sugges s ha
alo ising EU cheese whey pe mea e could inc ease LA and poly-LA p oduc ion by 40 and 125 imes, espec i ely. This could also lead o he p oduc ion o 4,000
k on p o ein- ich biomass, po en ially educing CO
2
emissions linked o EU ood and eed p oduc ion by 4.87% o 2.77% espec i ely.
1. In oduc ion
Lac ic acid (LA) is a biobased pla o m chemical wi h he second
la ges ma ke olume a e e hanol. The es ima ed annual LA ma ke
olume is 1,350 k on, wi h an expec ed annual g ow h o 8–16%
(Dusselie e al., 2013; Ma e al., 2023). LA is a e sa ile bulk chemical
ha can be con e ed in o a wide ange o p oduc s, including esins,
an i eeze, and la ou ing compounds (Dusselie e al., 2013). Fu he -
mo e, LA can be used as a subs a e o o he biological p ocesses p o-
ducing mo e aluable p oduc s such as mic obial p o ein (
€
6,860 pe
onne d y biomass) (Saka ika e al., 2022; Van Pe eghem e al., 2022).
This p omising pla o m chemical can be con e ed o he
biopolyme poly-lac ic acid (PLA). Depending on he op ical pu i y
(enan iome ic LA o ms L(+) o D(−)), di e en biopolyme s can be
p oduced wi h a ious mechanical p ope ies (Nase e al., 2021). L-LA
is p e e ed in he pha maceu ical and ood indus ies due o i s
enhanced assimila ion in he human body compa ed o D-LA
(Abdel-Rahman and Sonomo o, 2016; Rawoo e al., 2021). Lac ic acid
bac e ia (LAB), he key bioca alys s in LA p oduc ion, can u ilize a b oad
ange o subs a es, including indus ial side s eams, ood was e, and
lignocellulosic ma e ials (Abdel-Rahman and Sonomo o, 2016). Addi-
ionally, aluable by-p oduc s such as bac e iocins, exopolysaccha ides,
poly-β-hyd oxybu y a e, and p obio ic cul u es a e p oduced du ing LA
e men a ion (Mazzoli e al., 2014). The b oad ange o LA applica ions,
along wi h he di e se p oduc s p oduced by LAB, make hem ideal
candida es o bio e ine y concep s.
The economic easibili y o LA applica ions depends la gely on he
cos o he aw ma e ials and he equi emen o axenic condi ions (Li
* Co esponding au ho . CRETUS, Depa men o Chemical Enginee ing, Uni e sidade de San iago de Compos ela, 15075 San iago de Compos ela, Spain
E-mail add ess: [email p o ec ed] (A. Reguei a).
Con en s lis s a ailable a ScienceDi ec
Jou nal o En i onmen al Managemen
jou nal homepage: www.else ie .com/loca e/jen man
h ps://doi.o g/10.1016/j.jen man.2024.123529
Recei ed 12 July 2024; Recei ed in e ised o m 20 No embe 2024; Accep ed 27 No embe 2024
Jou nal o En i onmen al Managemen 373 (2025) 123529
A ailable online 3 Decembe 2024
0301-4797/© 2024 The Au ho s. Published by Else ie L d. This is an open access a icle unde he CC BY-NC-ND license (
h p://c ea i ecommons.o g/licenses/by-
nc-nd/4.0/ ).
e al., 2014). On an indus ial scale, LA p oduc ion u ilizes ood-g ade
eeds ocks such as suga bee , co n, and whea esul ing in compe i-
ion wi h ood esou ces (Spek eijse e al., 2019). Those ood-g ade
eeds ocks can accoun o up o 40–70% o he LA p oduc ion cos
(Ha is e al., 2024; Huang e al., 2023). In addi ion, indus ial LA p o-
duc ion elies on pu e cul u es equi ing s e ilisa ion, which inc eases
ope a ional expendi u e (OPEX) (Peinemann e al., 2019) and capi al
expendi u e (CAPEX) compa ed o non-axenic e men a ion (Chen and
Wan, 2017; L´
opez-G´
omez e al., 2019). In addi ion o high subs a e and
s e ilisa ion cos s, he p oduc ion is o en pe o med in ba ch p ocesses
(A idan and G eene , 2021; Mo issey e al., 2016), esul ing in low
p oduc i i ies (0.93–4.32 g/L/h), compa ed o chemical p oduc ion
he eby inc easing he o e all p oduc ion cos s (Do an, 2013).
La es esea ch aims o educe hose cos s by u ilizing mixed com-
muni ies and low- alue eeds ocks such as cheese whey pe mea e (CWP)
(Huang e al., 2023; Schü e le e al., 2024). This indus ial side s eam
has limi ed applica ions in he ood and eed sec o (e.g. bulk swee ene
ing edien in bake y p oduc s) and, i no alo ised, equi es cos ly
disposal ea men s (Khez i e al., 2016; Zo a e al., 2020). In e es -
ingly, CWP is one o he la ges dai y s eams, nex o milk, gi ing i
signi ican po en ial o be used as a eeds ock o pla o m chemical
p oduc ion. In 2020, he Eu opean Union (EU) p oduced 47 million ons
o CWP (Eu os a , 2020), which is ich in lac ose (50–60 g/L) (Byland,
2003; P aze es e al., 2012). Di e en s udies conduc ed in con inuous
s i ed ank eac o s (CSTR) ha e shown ha selec i e LA p oduc ion
can be achie ed by adjus ing speci ic ope a ional pa ame e s (Sakai and
Ezaki, 2006; Wu e al., 2015; Yang e al., 2022). Hyd aulic e en ion
ime (HRT) o less han 12 h has p o en o be an e ec i e selec ion ool
o LAB wi h high speci ic g ow h a es, he eby enhancing o e all LA
selec i i y (Choi e al., 2016; Saka ika e al., 2022). To u he minimize
he in asion o non-LAB, p omo e he g ow h o LAB, and enhance LA
selec i i y, o he s ingen condi ions such as acidic pH (<5.0) and
he mophilic empe a u es (50–60 ◦C) can be applied (Akao e al., 2007;
I oh e al., 2012; D.-H. Kim e al., 2012). Consequen ly, hese ope a-
ional pa ame e s (i.e. empe a u e, pH, and HRT) could lead o LAB
dominance, enabling selec i e LA p oduc ion while omi ing
s e ilisa ion.
P e ious esea ch on non-axenic LA e men a ion o side s eams has
o en ocused on op imizing a single pa ame e , usually LA selec i i y,
while o he c ucial ac o s like p oduc i i y, yield, concen a ion, and
op ical pu i y a e o en unde s udied. These pa ame e s a e essen ial o
he economic easibili y and scalabili y o LA applica ions. Addi ionally,
indus ial p ocesses end o ely on ood-g ade eeds ocks and a e
ene gy-in ensi e, posing sus ainabili y challenges. To add ess hese
limi a ions, his s udy aims o op imize mul iple ope a ional eac o
pa ame e s ( empe a u e: 45–55 ◦C, pH: 5.0–6.5, HRT: 10-1.5h) o
signi ican ly enhance key LA p oduc ion me ics: p oduc i i y, yield,
concen a ion, and op ical pu i y. Such a comp ehensi e explo a ion o
he combined e ec s o HRT, empe a u e, and pH on LA selec i i y has
no been p e iously conduc ed, making his s udy c ucial o b idging
he knowledge gap in selec i e LA p oduc ion. By simul aneously u i-
lizing one he la ges indus ial dai y side s eams, CWP, and pe o ming
non-axenic e men a ion, we aim no only o op imize LA p oduc ion
me ics bu also con ibu e o a mo e sus ainable and esou ce-e icien
LA p oduc ion p ocess. In addi ion, we assessed he mic obial commu-
ni y o LAB abundance, wi h a keen ocus on hei po en ial use as
p obio ic o p o ein ing edien in ood o eed applica ions. By iden i-
ying ope a ional condi ions ha ensu e bo h high LAB abundance and
mic obial sa e y, we no only imp o e LA p oduc ion bu also c ea e new
oppo uni ies o alo ising CWP. This s udy o e s a combined s a egy
o imp o ing LA p oduc ion while explo ing he use o LAB p oduced
unde non-axenic condi ions as ood o eed ing edien s.
2. Ma e ials and me hods
2.1. Cheese whey pe mea e cha ac e iza ion and nu ien
supplemen a ion
All expe imen s we e conduc ed using CWP collec ed om Milcobel,
a dai y p oduc manu ac u e in Langema k-Poelkapelle, Belgium. In
his acili y, cheese whey de i ed om mozza ella and chedda cheese
p oduc ion was combined, and he whey p o eins we e sepa a ed ia
ul a il a ion. The emaining CWP, con aining a limi ed amoun o
p o ein (0.30 ±0.09%), was collec ed in ba ches, combined, and s o ed
a −20 ◦C un il use. Mos LAB a e auxo ophic o amino acids
(Abdel-Rahman e al., 2013) and equi e o ganic ni ogen supplemen-
a ion (e.g., p o ein) o suppo hei g ow h. To de e mine he app o-
p ia e amoun o supplemen ed o ganic ni ogen needed o LAB g ow h
and high LA yields, a p elimina y expe imen was pe o med. The
o ganic ni ogen used in his expe imen was yeas ex ac (YE) due o i s
u ilisa ion on an indus ial scale (P ous e al., 2019). YE concen a ions
o 6 g/L, 9 g/L, and 12 g/L we e e alua ed in a CSTR, a ele an con-
di ions o his s udy (45 ◦C, pH 5.5, and HRT 10h). Amongs he YE
concen a ions e alua ed, 9 g/L was he lowes concen a ion ha did
no nega i ely impac he LA p oduc ion (Fig. S1). The e o e, 9 g/L o YE
was added o he CWP be o e he subsequen LA e men a ions. Cha -
ac e iza ion o CWP (Table 1), wi h and wi hou yeas ex ac supple-
men a ion, was pe o med by analysing concen a ions o ca bohyd a es
(lac ose, galac ose, glucose), o ganic acids (LA, ace ic acid, o mic acid,
p opionic acid, and bu y ic acid), anion and ca ion (Cl
−
, NO
2
−
,NO
3
−
,
PO
4
3−
, SO
4
2−
, Na
+
, NH
4
+
, K
+
, Ca
2+
, and Mg
2+
), o al ni ogen (TN),
chemical oxygen demand (COD), o al suspended solids (TSS), ola ile
suspended solids (VSS), pH, and elec ical conduc i i y (EC). The
Table 1
Cha ac e iza ion o cheese whey pe mea e wi h and wi hou yeas ex ac sup-
plemen a ion. The alues p esen ed a e he a e age alues o all he di e en
cheese whey pe mea e ba ches used in his wo k (n ≥6).
Pa ame e Cheese whey
pe mea e
Cheese whey pe mea e
supplemen ed wi h 9 g/L yeas
ex ac
Uni
pH 6.08 ±0.09 6.19 ±0.09 /
EC 7.80 ±0.38 8.45 ±0.51 mS/
cm
TN 0.29 ±0.01 0.97 ±0.08 g/L
O ganic o al
ni ogen
0.18 ±0.01 0.80 ±0.02 g/L
COD 86.3 ±2.4 83.0 ±4.3 g/L
Ca bohyd a es
Lac ose 62.8 ±6.1 65.3 ±6.3 g/L
Glucose 0.18 ±0.10 0.21 ±0.11 g/L
Galac ose 1.48 ±0.40 1.79 ±0.27 g/L
O ganic acids
Lac ic acid 1.36 ±0.03 1.41 ±0.03 g/L
Ace ic acid 0.21 ±0.02 0.09 ±0.03 g/L
Bu y ic acid 0.61 ±0.05 0.28 ±0.04 g/L
P opionic acid ND 2.24 ±1.14 mg/L
Ci ic acid 2.11 ±0.08 9.07 ±0.25 mg/L
Anions
Cl
−
1,390 ±57 1,545 ±45 mg/L
NO
2
−
ND ND mg/L
NO
3
−
26.3 ±2.7 29.3 ±1.5 mg/L
PO
4
3-
1,473 ±27 1,856 ±15 mg/L
SO
4
2-
195 ±12 264 ±23 mg/L
Ca ions
Na
+
535 ±3 527 ±2 mg/L
NH
4
+
138 ±24 168 ±10 mg/L
K
+
2,280 ±57 2,667 ±40 mg/L
Ca
2+
417 ±5 387 ±9 mg/L
Mg
2+
113 ±5 109 ±2 mg/L
TSS/VSS
TSS 1.2 ±0.3 1.6 ±0.3 g/L
VSS 0.9 ±0.3 1.1 ±0.3 g/L
ND: no de ec ed.
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
2
speci ic analy ical echniques a e desc ibed in de ail in sec ion 2.4.
2.2. Inoculum
Cheese whey, p io o e e se osmosis and ul a il a ion, was
collec ed om Milcobel and s o ed a −20 ◦C. The cheese whey was used
as inoculum o all he eac o uns. To ensu e high mic obial ac i i y o
he inoculum, i was incuba ed be o e eac o inocula ion o 48 h a
45 ◦C, 50 ◦C, and 55 ◦C, depending on he espec i e LA e men a ion in
a CSTR.
2.3. Con inuous e men a ion o cheese whey pe mea e o lac ic acid
p oduc ion
The non-axenic e men a ion o CWP o LA was pe o med in a
double jacke ed glass CSTR wi h a o al olume o 0.6 L and a wo king
olume o 0.2 L, magne ically s i ed a 400 pm. A o al o se en eac o
uns, wi h a ying empe a u e and pH we e conduc ed unde anae obic
condi ions, which we e ensu ed by pu ging he eac o wi h N
2
p io o
inocula ion and main ained using a ubbe seal. Each eac o un e al-
ua ed ei he a speci ic empe a u e (45 ◦C, 50 ◦C, o 55 ◦C) o pH alue
(5.0, 5.5, 6.0, o 6.5). Fi s , he op imal empe a u e (45 ◦C, 50 ◦C, o
55 ◦C) o LA p oduc i i y, selec i i y, op ical pu i y, and yield was
de e mined a a ixed pH o 5.5. The pH was measu ed wi h a pH elec-
ode (662–1774, VWR®) and was con olled a he a ge ed alue
h ough addi ion o 2 M NaOH using a Mic odos MP pH con olle
(Ve de lex VP). A e es ablishing he op imal empe a u e o he LA
e men a ion pa ame e s, di e en pH alues (5.0, 5.5, 6.0, and 6.5)
we e e alua ed a he op imal empe a u e. Each un began wi h a
eshly incuba ed inoculum (10 / %), p epa ed by incuba ing un il-
e ed cheese whey a he speci ic eac o empe a u e (sec ion 2.2). A e
inocula ion, he eac o was ope a ed in ba ch mode o 48 h, ollowed
by con inuous ope a ion a an HRT o 10h. Once s eady s a e was
es ablished, lowe HRTs we e g adually e alua ed (10 h - 6 h - 3 h - 2.25
h - 1.5 h). S eady s a e ope a ion was de ined as a maximum di e ence
o ±5% be ween LA concen a ion alues o e a leas 10 HRTs. Reac o
uns we e immedia ely s opped i eac o ailu e occu ed. The ba ch
mode and subsequen con inuous ope a ion we e pe o med wi h CWP,
supplemen ed wi h 9 g/L YE (Table 1), wi h a concen a ion o 63 ±6 g/
L lac ose, 0.2 ±0.1 g/L glucose, and 1.5 ±0.4 g/L galac ose. A 15 mL
sample was aken e e y 1–2 HRT and analysed o o ganic acids, pH, and
elec ic conduc i i y. The emaining sample was s o ed a −20 ◦C. Fo
s eady s a e samples, addi ional ope a ional pa ame e s we e measu ed
including C2-C8 a y acids, ca bohyd a es, and e hanol. Fo mic obial
communi y composi ion analysis, a 2 mL sample was cen i uged a
19,100 g o 5 min (Eppendo ™ Cen i uge 5430), he pelle was s o ed
a −20 ◦C un il analysis (sec ion 2.5).
2.4. Analy ical echniques
O ganic acids such as LA, ace ic acid, o mic acid, p opionic acid,
and bu y ic acid we e analysed wi h a 930 Compac Ion Ch oma og aph
Flex (Me ohm®, Swi ze land), using a Me osep O ganic acids 250/7.8
column wi h inline bica bona e emo al. The column was equipped wi h
a Me osep O ganic acid Gua d/4.6. Ca ions we e analysed using a 761
Compac Ion Ch oma og aph (Me ohm®, Swi ze land). The ion ch o-
ma og aph was equipped wi h a Me osep C6-250/4.0 column and a
conduc i i y de ec o . Anions such as chlo ide, ni a e, sulpha e we e
analysed using a Me h om 930 compac Ion Ch oma og aph, equipped
wi h a conduc i i y de ec o and a Me osep A supp 5–150/4.0 column
(Me ohm®, Swi ze land). C2-C8 a y acids we e analysed (including
iso o ms C4-C6) as desc ibed in Ande sen e al. (2014). The op ical D-LA
and L-LA pu i y was measu ed using an enzyma ic L-LA assay ki
(Megazyme, Uni ed Kingdom). Lac ose, galac ose, glucose, and e hanol
concen a ions we e de e mined by high-pe o mance liquid ch oma-
og aphy (Shimadzu LC-2030C Plus P ominence-i-se ies) equipped wi h
a e ac i e index de ec o (RID-20A), se a 40 ◦C, and an Aminex
HPX-87H (Bio ad) column p o ec ed wi h a mic o gua d ca idge. A
mobile phase o 5 mM H
2
SO
4
wi h 1% ace oni ile a a low a e o 0.5
mL/min was used. The column empe a u e was kep a 30 ◦C. De e -
mina ion o COD and TN we e pe o med using comme cial ki s
(NANOCOLOR®; MACHEREY-NAGEL GmbH & Co. KG). TSS and VSS
we e analysed acco ding o S anda d Me hods (APHA, 2017). Finally,
pH and elec ical conduc i i y we e measu ed wi h a pH elec ode
(SP10B, Conso ) and EC elec ode (SK21, Conso ).
2.5. Molecula echniques
The mic obial communi y composi ion was analysed o e e y con-
di ion ( empe a u e and pH) a HRT 6 h and 1.5 h, excep o 55 ◦C due
o eac o ailu e a HRT 10 h. The mic obial biomass was dis up ed by
bead bea ing wi h a Powe Lyze (Qiagen, Venlo, he Ne he lands).
Subsequen ly, DNA was ex ac ed wi h phenol/chlo o o m
(Vilchez-Va gas e al., 2013). A 10
μ
L aliquo was sen ou o LGC ge-
nomics GmbH (Be lin, Ge many) o lib a y p epa a ion and sequencing
on an Illumina Miseq pla o m wi h 3 chemis y. Fo ampli ica ion o
he DNA ex ac , ollowing p ime s we e used 341F (5
′
-CCT ACG GGN
GGC WGC AG −3
′
) and 785Rmod (5
′
-GAC TAC HVG GGT ATC TAA
KCC-3
′
). Assembly and clean-up o he eads we e based on he MiSeq
SOP desc ibed by he Schloss lab. DADA2 ( . 1.16) was used o assemble
he eads in o con igs, pe o m alignmen -based quali y il e ing and
emo e chime as (Callahan e al., 2016). Taxonomy assignmen was
pe o med using a naï e Bayesian classi ie and SILVA NR 132 clus-
e ing con igs in o amplicon sequence a ian s (ASV).
2.6. Calcula ions
2.6.1. Lac ic acid p oduc i i y
The LA p oduc i i y (P
LA
) was calcula ed as he LA concen a ion
(C
LA
) di ided by he HRT (Eq. 1.)
LA p oduc i i y(g
L⋅h)=
CLA (g
L)
HRT (h)(1)
2.6.2. Lac ic acid selec i i y
The LA selec i i y (S
LA
) was calcula ed as he C
LA
di ided by he
concen a ion o all o ganic acids and e hanol p oduced (Eq. 2.)
LA selec i i y (%) =
CLA (g
L)
∑CP oduc s(g
L)×100 (2)
2.6.3. Lac ic acid yield
The LA yield (Y
LA
) was calcula ed as he C
LA
di ided by he con-
cen a ion o ca bohyd a es (glucose, galac ose, and lac ose) p esen in
he CWP be o e e men a ion (Eq. 3.)
LA yield (%) =
CLA (g
L)
∑Cin luen ca bohyd a es (g
L)×100 (3)
2.6.4. L(+)- lac ic acid pu i y/D(−)-lac ic acid pu i y
The op ical pu i y o L-LA (OP
L-LA
) was de ined as he concen a ion
o L-LA concen a ion measu ed di ided by he o al concen a ion o LA
(Eq. (4)). Fo he op ical pu i y o D-LA (OP
D-LA
), he o al concen a ion
o LA was sub ac ed om he L-LA concen a ion and di ided by he
o al concen a ion o LA (Eq. 5.).
L−LA op ical pu i y(%) =
CL−LA (g
L)
CLA (g
L)×100 (4)
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
3
D−LA op ical pu i y(%) =
CLA(g
L)−CL−LA (g
L)
CLA (g
L)×100 (5)
3. Resul s & discussion
3.1. S ee ing a non-axenic selec i e lac ic acid e men a ion h ough
p ocess pa ame e manipula ion
3.1.1. Lowe ing eac o empe a u e om 55 ◦C o 45 ◦C esul ed in
enhanced lac ic acid p oduc ion
To assess he e ec o empe a u e (45–55 ◦C) on LA e men a ion,
he pH was con olled a 5.5 while he HRT was g adually dec eased
om 10 h o 1.5 h. A all e alua ed empe a u es (45 ◦C, 50 ◦C, and
55 ◦C), LA was he main p oduc a concen a ions o 9.0–47.4 g/L (Fig
S2) ollowed by ace ic acid a a ound 1.0 g/L, ega dless o he condi-
ion. Low concen a ions o o he by-p oduc s such as bu y ic acid (up o
1.4 g/L), and o mic acid (up o 0.5 g/L) we e p oduced while e hanol o
p opionic acid we e no de ec ed. The highes LA selec i i y in his s udy
(97.8%) was ound a 45 ◦C (Table 2). The cu en wo k achie ed highe
LA selec i i y han p e iously epo ed in li e a u e (5.0–89.0%) o LA
e men a ions a 45 ◦C (D.-H. Kim e al., 2012; Saka ika e al., 2022). We
achie ed simila LA selec i i y (95.8%) a 50 ◦C (Table 2), which aligns
wi h he indings o Kim e al. (2012), who achie ed a LA selec i i y o
97.0% a 50 ◦C. Howe e , a 55 ◦C, we ob ained a LA selec i i y o only
80.7% (Table 2) which di e s om esul s ob ained in p e ious s udies
ha epo ed a 100% LA selec i i y a his empe a u e (D.-H. Kim e al.,
2012). These s udies ocused mainly on empe a u es anging om 50 o
55 ◦C based on he hypo hesis ha highe empe a u es enable selec i e
LA p oduc ion due o he s ingen condi ions (Akao e al., 2007; Choi
e al., 2016; Yang e al., 2022). Howe e , he LA selec i i y we ob ained
a 55 ◦C was he lowes o all e alua ed empe a u es.
A 45 ◦C, he LA p oduc i i y anged be ween 5.0 and 12.3 g/L/h
depending on he HRT (Fig. 1A). When he empe a u e was 50 ◦C, an
o e all lowe LA p oduc i i y o 3.7–9.6 g/L/h was obse ed. A 55 ◦C
he LA p oduc i i y was conside ably lowe (0.9 g/L/h) a an HRT o 10
h, compa ed o he o he empe a u es (Fig. 1A). Hence, lowe HRT
alues we e no e alua ed a his empe a u e. P e ious non-axenic LA
e men a ion pe o med a 50 ◦C only achie ed a LA p oduc i i y o
0.1–1.8 g/L/h a 50 ◦C (D.-H. Kim e al., 2012; Schü e le e al., 2024;
Yang e al., 2022) which is conside ably lowe han he esul s ob ained
in ou s udy a 50 ◦C (3.7–9.6 g/L/h). Non-axenic LA e men a ions
pe o med a 45 ◦C achie ed a LA p oduc i i y o 2.0–9.9 g/L/h a a HRT
o 1.5–12 h (Saka ika e al., 2022; Schü e le e al., 2024), which is in
line wi h he esul s ob ained he e.
The LA yield and concen a ion we e s ongly impac ed by he
empe a u e. Depending on he HRT, expe imen s a 45 ◦C and 50 ◦C
achie ed LA yields o 24.5–100.9% and 23.9–72.9% (Fig. 1B) and con-
cen a ions o 15.3–47.4 g/L and 13.1–36.8 g/L (Fig. S2). The lowes LA
yield o 15.6% was ob ained a 55 ◦C, co esponding o a concen a ion
o 9.0 g/L. P e ious li e a u e epo s eached a compa able maximum
LA yield (92.0%) a 45–50 ◦C (Choi e al., 2016; Saka ika e al., 2022).
Ne e heless, o he s udies a 55 ◦C achie ed highe LA yields (43–92%)
(D.-H. Kim e al., 2012; Yang e al., 2022).
In summa y, we obse ed imp o ed LA e men a ion pe o mance as
he empe a u e was dec eased om 55 ◦C o 45 ◦C. This inc ease can be
a ibu ed o he g ow h kine ics and yields cha ac e is ic o he mo-
philic LAB species, which a e ypically isola ed om dai y side s eams
(Ma asco e al., 2022). These species, such as Lac obacillus delb ueckii
and S ep ococcus he mophilus, demons a ed a 147–170% inc ease in
g ow h a e and a 170–194% inc ease in biomass yield wi h a empe -
a u e dec ease om 50 ◦C o 45 ◦C (Adambe g e al., 2003). Mo eo e ,
no g ow h was obse ed o ei he species a 55 ◦C. Howe e , o he
he mophilic LAB, such as Bacillus coagulans can be en iched a 55 ◦C
(Choi e al., 2016). None heless, hose LAB species a ou ing empe a-
u es o 55 ◦C, o highe , ha e lowe g ow h a es and biomass yields
compa ed o S ep ococci and Lac obacilli (Bo ja e al., 1995; Glase and
Venus, 2017). I is highly likely ha he mic obial communi y es ab-
lished in ou s udy a 55 ◦C had signi ican ly lowe biomass yield and
g ow h a e compa ed o ou mic obial communi y a 45 ◦C. In combi-
na ion wi h he sho e HRT (≤10h) applied in ou s udy, compa ed o
o he s udies (≥12h), his likely esul ed in he washou o he mophilic
LAB biomass and, consequen ly, educed he o e all LA e men a ion
pe o mance. This could explain he disc epancy be ween he esul s we
ob ained a 55 ◦C and o he s udies.
Cu en s a e o he a o en e ol es a ound maximizing one o wo
Table 2
P oduc selec i i y ob ained in all e alua ed HRTs du ing con inuous lac ic acid
e men a ion a pH 5.5 and empe a u es o 45 ◦C, 50 ◦C and 55 ◦C. Mean alues
and s anda d de ia ion a e p esen ed (n =3). Bo h e hanol and p opionic acid
we e no de ec ed.
Tempe a u e
(◦C)
Lac ic acid
(%)
Ace ic acid
(%)
Fo mic acid
(%)
Bu y ic acid
(%)
45 97.8 ±0.4 2.1 ±0.2 ND ND
50 95.8 ±1.7 1.9 ±0.8 2.2 ±0.9 ND
55 80.7 ±2.0 10.9 ±2.4 0.8 ±0.7 7.6 ±0.4
ND: no de ec ed, hence no calcula ed.
Fig. 1. The e ec o empe a u e (45, 50, and 55 ◦C) a pH 5.5 on (A) lac ic acid
p oduc i i y and (B) lac ic acid yield. A e age alues (n ≥3) ±s anda d de-
ia ion a s eady s a e in each hyd aulic e en ion ime (HRT) a e p esen ed.
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
4
LA e men a ion pa ame e s. Fo example, indus ial LA p oduc ion in
axenic ba ch eac o s ocuses on ob aining high LA yield and concen-
a ions bu o en su e s om low p oduc i i y. The u ilisa ion o mixed
communi ies has demons a ed high LA selec i i y bu o en lacks p o-
duc i i y, i es, o yields. Ou esul s showed highe LA p oduc i i y,
i es, and selec i i y (P
LA
: 12.3 g/L/h, C
LA
: 28.1 g/L, S
LA
: 97.6%)
compa ed o p e iously epo ed maximum LA p oduc i i ies o pu e
cul u es (P
LA
: 11.2 g/L/h, C
LA
: 22.4 g/L, S
LA
: 100%) and mixed com-
muni ies (P
LA
: 9.9 g/L/h, C
LA
: 14.9 g/L, S
LA
: 89%) achie ed in a CSTR
(G¨
oksungu and Gü enç, 1997; Saka ika e al., 2022). Howe e , as seen
in ou s udy and o he s, lowe LA yields (43.0–45.4%) we e ob ained
while maximizing LA p oduc i i y. This ade-o be ween yield and
p oduc i i y is a common challenge in CSTR ope a ions. Despi e his,
ou s udy demons a ed signi ican imp o emen s when maximizing LA
yield. A an HRT o 10h (45 ◦C, pH 5.5) we ob ained a LA yield o 100%
(P
LA
: 5.0 g/L/h, C
LA
: 47 g/L, S
LA
: 97.9%). In con as , a pu e cul u e LA
e men a ion (G¨
oksungu and Gü enç, 1997) only achie ed a 76.8% LA
yield (P
LA
: 4.0 g/L/h, C
LA
: 40 g/L, S
LA
: 100%) and o mixed communi y
(Saka ika e al., 2022) a LA yield o 75.4% was achie ed (P
LA
: 3.9 g/L/h,
C
LA
: 23.4 g/L, S
LA
: 74%). No ably, ou non-axenic LA e men a ion
p ocess can no only be maximized o LA p oduc i i y bu also o LA
yield while ou pe o ming p e ious s udies in bo h condi ions.
3.1.2. Ci cumneu al pH can signi ican ly imp o e lac ic acid e men a ion
pe o mance
Fou con inuous LA e men a ions we e con olled a pH alues o
5.0, 5.5, 6.0, and 6.5 a 45 ◦C, which was he empe a u e ha esul ed
in he bes pe o mance in his wo k (sec ion 3.1.1). In each eac o un
he HRT was g adually dec eased om 10 h ill 1.5 h. We achie ed high
LA selec i i y (≥95.0%) in non-axenic LA e men a ion independen o
he ope a ional pH. A all es ed pH alues LA was he main p oduc
(94.9–97.8%) ollowed by ace ic acid and o mic acid wi h a selec i i y
o 2.0–5.1% and 0.7–1.5%, espec i ely (Table 3). In con as , p e ious
s udies e alua ing highe pH alues (6.0–6.5) epo ed a signi ican
dec ease in LA selec i i y. Speci ically, inc easing he pH om 3.5 o 5.5
o 6.5, independen o he empe a u e, dec eased he LA selec i i y
om 73-92% o 25–38% (I oh e al., 2012; Yang e al., 2022).
Besides, inc easing he pH om 5.0 o 6.5 esul ed in a signi ican
inc ease in LA p oduc i i y om 3.1 o 7.6 g/L/h o 5.5–27.4 g/L/h
(Fig. 2A). To he bes o ou knowledge, he alue o 27.4 g/L/h ob ained
a pH 6.5 and HRT 1.5 h is he highes LA p oduc i i y epo ed in axenic
and non-axenic LA e men a ion in a CSTR. In conjunc ion wi h an
inc eased LA p oduc i i y, a posi i e e ec on he LA yield and con-
cen a ion was obse ed inc easing he LA yield om 16.1 o 47.3% a
pH 5.0 o 70.0–99.0% a pH 6.5 (Fig. 2B). A simila obse a ion was
made o he LA concen a ion which inc eased om 13.0 o 29.7 g/L a
pH 5.0 o 44.2–53.1 g/L a pH 6.5 (Fig S3). Those indings a e in line
wi h p e ious s udies as inc easing he pH esul ed in an LA p oduc i i y
and yield inc ease. Howe e , he e an inc ease in pH esul ed in o e all
lowe LA selec i i y which is no obse ed in ou s udy (I oh e al., 2012;
Yang e al., 2022).
O e all, he lowe LA p oduc i i y and LA yield ob ained a lowe pH
alues is mos likely caused by he inc eased ac ion o undissocia ed LA
which can cause mic obial g ow h inhibi ion, he eby esul ing in lowe
biomass yields and, a some poin , subs a e u ilisa ion (I oh e al., 2012;
Yang e al., 2022). Howe e , in ou s udy inc easing he pH did no al e
he LA selec i i y which was obse ed in o he s udies (Yang e al.,
2022). This di e ence is mos ly likely due he HRT di e ence in ou
s udy (≤10h) and o he s udies (≥12h). A sho e HRT can apply a se-
lec ion p essu e owa ds as g owing o ganisms such as LAB ou -
compe ing o he o ganisms esponsible o by-p oduc o ma ion
(Rombou s e al., 2020).
He e, we achie ed high LA selec i i y (≥95.0%) unde non-axenic
LA e men a ion independen om he ope a ional pH. Addi ionally,
inc easing he pH om 5.0 o 6.5 showed an inc ease in LA yield
(99.0%), concen a ion (53.1 g/L), and we eached he highes epo ed
LA p oduc i i y o 27.4 g/L/h ill now. I su passed p e ious eco ds o
axenic (11.2 g/L/h) and non-axenic (9.9 g/L/h) LA e men a ions, pe -
o med a 45 ◦C, wi h a ac o o wo o highe (G¨
oksungu and Gü enç,
1997; Saka ika e al., 2022). Bo h p e ious p oduc i i y eco ds only
achie ed a LA yield o a ound 50.0% while he LA e men a ion we
pe o med esul ed in a LA yield o 70.0%. These esul s demons a e
ha non-axenic ope a ions can compe e and e en ou pe o m axenic LA
e men a ions and simul aneously educing he OPEX by 15% and
CAPEX by 10%, compa ed o axenic ope a ions. This can gi e ise o
mo e LA applica ions becoming economically easible (Chen and Wan,
2017; L´
opez-G´
omez e al., 2019).
Table 3
A e age p oduc selec i i y o (n =3) o con inuous lac ic acid e men a ions a
ixed empe a u e 45 ◦C and di e en pH alues 5.0–6.5. Bo h e hanol and
p opionic acid we e no de ec ed.
pH Lac ic acid (%) Ace ic acid (%) Fo mic acid (%) Bu y ic acid (%)
5.0 94.9 ±2.7 5.1 ±2.8 ND ND
5.5 97.8 ±0.4 2.1 ±0.2 ND ND
6.0 97.3 ±0.8 2.0 ±0.4 0.7 ±0.7 ND
6.5 96.2 ±0.9 2.3 ±0.7 1.5 ±0.2 ND
ND: no de ec ed, hence no calcula ed.
Fig. 2. The e ec o pH alues 5.0 ill 6.5 a 45 ◦C on (A) lac ic acid p oduc-
i i y (B) lac ic acid yield. A e age alues (n ≥3) ±s anda d de ia ion a
s eady s a e in each HRT a e p esen ed.
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
5
3.1.3. A low HRT is c ucial o achie e high lac ic acid selec i i y
F om he p e ious eac o expe imen s e alua ing bo h pH (5.0–6.5)
and empe a u e (45–55 ◦C), we can also in e he e ec o he HRT on
he e men a ion pe o mance. The e alua ed HRTs did no signi ican ly
impac he LA selec i i y which anged be ween 92.3 and 98.3%
(Table S2). This is mos likely a consequence o LAB ha ing a kine ic
ad an age a he sho HRTs applied in ou expe imen s o e o he
mic obial g oups (Rombou s e al., 2020). Di e en con inuous
non-axenic e men a ions wi h a HRT highe han 12h esul ed in he
co-gene a ion o o he p oduc s such as bu y ic acid and ace ic acid a
45 ◦C (Saka ika e al., 2022) and 55 ◦C (Choi e al., 2016). In bo h cases,
educing he HRT om 48-120h o 6–12 h inc eased he LA selec i i y
om 6% up o 95% (Choi e al., 2016; Saka ika e al., 2022).
As commonly obse ed in CSTR ope a ions, dec easing he HRT had
a conside able posi i e e ec on he LA p oduc i i y in ou expe imen s.
In all e alua ed condi ions dec easing he HRT om 10 h o 1.5 h
(Figs. 1A and 2A) esul ed in a LA p oduc i i y inc ease om 3.1 o 5.5
g/L/h o 7.6–27.4 g/L/h. Howe e , dec easing he HRT nega i ely
a ec ed he LA yield, educing i om 44.9-99.0% o 21.2–70.0%
(Figs. 1B and 2B). Thus a clea ade-o exis s be ween LA p oduc i i y
and LA yield, which is also commonly obse ed in li e a u e
(Abdel-Rahman e al., 2016; Choi e al., 2016; D.-H. Kim e al., 2012).
Bo h LA p oduc i i y and LA yield should be maximized as bo h can
educe cos s h ough educing eac o size and inc easing o e all
p oduc yields (Do an, 2013).
3.1.4. Ci cumneu al pH o 6–6.5 a ou s he o ma ion o L-lac ic acid
o e D-lac ic acid
Highly op ically pu e LA is impo an o ce ain applica ions. Fo
example, op ically pu e L-LA is p e e ed o pha maceu ical applica-
ions and poly-L-LA p oduc ion while D-LA used o poly-D-LA p oduc-
ion. Indus ial p oduc ion o ei he L-LA o D-LA p oduc ion is
accomplished using speci ic o ganisms ha exclusi ely p oduce he
espec i e enan iome s. To his end, we analysed op ical pu i y du ing
s eady s a e ope a ions (Fig. 3). Inc easing he empe a u e om 45 ◦C
o 55 ◦C, a pH 5.5, inc eased he L-LA om 12.6 ±2.8% o 53.0% ±
6.5% (Fig. 3A). In con as , o he s udies pe o med a 55 ◦C a ained
highe op ical L-LA pu i y, eaching 96.7% (Akao e al., 2007; D.-H. Kim
e al., 2012; Yang e al., 2022). Addi ionally, hey ound ha in mos
cases Bacillus coagulans, a known L-LA p oduce , is he main con ibu o
o he high L-LA op ical pu i y (Akao e al., 2007; D.-H. Kim e al., 2012;
Yang e al., 2022). Howe e , as shown by Saka ika e al. (2022)
non-axenic eac o s ope a ed a lowe HRT (<6h) wash ou Bacillus
coagulans. Compa ed o o he s udies pe o ming a 55 ◦C we ope a ed a
much lowe HRT (≤10h), washing ou Bacillus coagulans, which is mos
likely he cause o he low L-LA op ical pu i y achie ed in ou s udy.
Addi ionally, a acidic pH (5.0–5.5) and 45 ◦C we achie ed high D-LA
op ical pu i y wi h a maximum o 91.3% while mo e neu al pH alues
(6.0–6.5) yielded a maximum L-LA op ical pu i y o 98.3% a 45 ◦C
(Fig. 3B). Yang e al. (2022) a ibu ed he shi in enan iome p oduc-
ion o a change in he communi y, which is also he cause in ou s udy,
as discussed in sec ion 3.2. Ne e heless, we ha e shown ha h ough
selec ion o eac o ope a ions high op ical pu i ies o L-LA o D-LA can
ob ained unde non-axenic condi ions. P o ided he ob ained pu i y is
su icien o pu i ica ion, we can omi he need o s e ilisa ion he eby
enhancing he economic easibili y o LA applica ions.
3.2. Inc easing empe a u e and pH can lead o a highly en iched
Lac obacilli communi y
The o iginal composi ion o he inoculum was domina ed by a
S ep ococcus species (ASV 4) (Fig. 4) and was subs an ially di e en
om he mic obial communi y obse ed du ing he di e en eac o
ope a ions. Ou esul s show ha he dominan gene a a e igh ly con-
nec ed wi h HRT, empe a u e, and pH, and, mo e impo an ly, ha
ema kably high en ichmen s o a single genus could be achie ed, which
is an impo an quali y o he alo isa ion o he p oduced biomass in
ood o eed applica ions.
A 45 ◦C and pH 5.0, Lac obacillus was he dominan genus wi h a
ela i e abundance o 74.3–90.3% composed o h ee di e en Lac o-
bacillus ASVs (Fig. 4). Addi ionally, ASVs belonging o he Rahnella,
Se a ia, and Lac ococcus gene a we e ound a a ela i e abundance o
4.0–19.8% depending on he HRT. Bo h Rahnella and Se a ia a e
acul a i e anae obic, non-lac ose e men ing oppo unis ic pa hogens
bu can p oduce LA, ace ic acid and e hanol om glucose (Ba man e al.,
2020; K. Y. Kim e al., 1997; Pa adh, 2015) while Lac obacillus and
Lac occocus a e bo h LAB (Onyeaka and Nwabo , 2022). Inc easing he
pH om 5.0 o 5.5 changed he ela i e abundance o Lac obacillus (ASV
1) om 30.5% o 93.5% while he o he gene a p esen a pH 5.0 no ed a
ela i e abundance below 2.0% a pH 5.5. A pH 6.0 and HRT 6 h,
Lac obacillus (ASV 1) achie ed a 99.2% ela i e abundance. Rema kably,
ASV 1, belonging o he genus Lac obacillus, was no de ec ed a pH 6.0,
HRT 1.5 h. A ha speci ic condi ion, he p e iously Lac obacillus-do-
mina ed communi y shi ed owa ds a S ep ococci-domina ed commu-
ni y, wi h a ela i e abundance o 98.8%. Simila ly, inc easing he pH o
6.5 esul ed in a ela i e abundance o 95.9–97.0% o he same h ee
S ep ococcus ASVs ound a pH 6.0 and an HRT o 1.5 h. Addi ionally,
aising he empe a u e om 45 ◦C o 50 ◦C a a ixed pH o 5.5 inc eased
he ela i e abundance o Lac obacillus (ASV 1) om 90.3-93.5% o
Fig. 3. The e ec o (A) empe a u e (45, 50, and 55 ◦C) a pH 5.5 and (B) pH
(5.0, 5.5, 6.0 and 6.5) a 45 ◦C on op ical L-LA pu i y. A e age alues (n ≥3) ±
s anda d de ia ion a s eady s a e in each HRT a e p esen ed.
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
6
97.0–98.6%.
The pH was he mos c i ical eac o pa ame e di e en ia ing be-
ween Lac obacillus (pH 5.0–5.5) and S ep ococcus (pH 6.0–6.5) com-
muni y. These obse a ions a e in line wi h he indings o Schü e le
e al. (2024). They obse ed simila communi y shi s, linked o he
ope a ional pH, in a non-axenic LA e men a ion s a ing om acid
cheese whey. A a pH o 5.5 (38–50 ◦C) hey no ed a 96.7–98.8% ela i e
abundance o Lac obacillus. Inc easing he pH o 6.5 a 44 ◦C dec eased
he ela i e abundance o Lac obacillus o 31.1% while S ep ococcus
became he dominan genus wi h a ela i e abundance o 65.5%
(Schü e le e al., 2024). This shi in communi y, caused by he pH
inc ease, can be linked o he op imal g ow h condi ions and kine ics o
hese gene a. The op imal g ow h empe a u e and pH o S ep ococci
a e app oxima ely 40–45 ◦C and wi hin he pH ange o 6.0–6.9 (Chen
e al., 2016). In con as , he op imal pH o Lac obacilli a ies be ween
4.5 and 6.0 (´
Sli˙
zewska and Chlebicz-W´
ojcik, 2020). Adambe g e al.
(2003) e alua ed he g ow h a e o di e en LAB a pH 6.5 (45 ◦C),
no ing ha S ep ococcus he mophilus had a g ow h a e o 2.25 h⁻
1
,
compa ed o 0.3–1.2 h⁻
1
o a ious Lac obacilli. This disc epancy likely
explains he dominance o S ep ococcus a ci cumneu al pH in ou
s udy, ou compe ing Lac obacilli.
O e all, he mic obial communi y, wi h a high abundance o S ep-
ococcus, showed an imp o ed e men a ion pe o mance (P
LA
: 27.4 g/L/
h, Y
LA
: 70.0%, C
LA
: 44.2 g/L, OP
L-LA
: 90.8%) compa ed o he Lac oba-
cillus-domina ed communi ies (P
LA
: 12.3 g/L/h, Y
LA
: 43.4%, C
LA
: 28.1 g/
L, OP
D-LA
: 93.2%). The di e ence in op ical pu i y be ween he di e en
communi ies is linked o he p esence o he ep esen a i e dominan
ASVs. A highly en iched Lac obacillus (ASV 1) communi y showed a high
D-LA op ical pu i y (84.0–93.2%). Acco ding o he basic local align-
men sea ch ool (BLAST), his ASV showed a 100% simila i y wi h
Lac obacillus delb ueckii, a known D-LA p oduce (Sahoo and Jaya aman,
2019). In con as , he high L-LA op ical pu i y (90.8–98.3%) was
a ibu ed o he exclusi e L-LA p oduc ion o all S ep ococcus sp. (Toi
e al., 2014). The inc eased e men a ion pe o mance a ci cumneu al
pH is linked o he speci ic lac a e p oduc ion cha ac e is ic o he
di e en gene a. As shown by Adambe g e al. (2003), S ep ococcus
species had a 2.8 imes highe speci ic lac a e p oduc ion a e (15.3 gLA.
gCDW
−1
. h
−1
) a pH 6.5 compa ed o Lac obacillus delb ueckii (5.4 gLA.
gCDW
−1
.h
−1
) a pH 5.5. This clea ly indica es ha he highe LA p o-
duc i i y ob ained in ou s udy a e linked o he highe speci ic LA
p oduc ion a e a ibu ed o S ep ococci. No ably, nei he S ep ococci
(ASV1-3) and Lac obacillus (ASV1) we e de ec ed in he ini ial inoculum,
ye hey eme ged as he dominan ASVs due he selec i e in luence o he
eac o ope a ional pa ame e s. A s udy pe o med by Fo es e al.
(2012) obse ed ha independen o he inoculum sou ce simila
p oduc spec a and e men a ion pe o mance can be ob ained. Ou
s udy highligh s, ha non-axenic e men a ions can be s ee ed owa ds
selec i e LA p oduc ion (Choi e al., 2016; Schü e le e al., 2024; Yang
e al., 2022) wi h high D-LA o L-LA op ical pu i y (Akao e al., 2007;
Sakai and Ezaki, 2006) due o selec ion o ope a ional eac o ope a-
ional pa ame e s.
Besides LA p oduc ion, he p oduced biomass can be o in e es o
he ood and eed sec o as a p o ein sou ce o as p obio ic. Howe e , as
we ha e an open e men a ion he communi y composi ion is a de e -
mining ac o o ob ain a sa e p oduc . The sa e y o o ganisms can be
assessed h ough se e al isk assessmen s. Howe e , some o ganisms
al eady ecei ed a Quali ied P esump ion o Sa e y (QPS) s a u e by he
EFSA as hey do no aise any sa e y conce ns o humans, animals and
he en i onmen (EFSA e al., 2022). We achie ed ela i e abundances
up o 99.2% o L. delb ueckii, indica ing ha al hough he eac o was
ope a ed in non-axenic condi ions, we could ob ain a highly en iched
communi y o one species wi hou losing pe o mance o eac o s a-
bili y. E en hough high ela i e abundance does no ensu e he sa e y o
he p oduc ou indings imply he po en ial u ilisa ion o
well-moni o ed non-axenic p ocesses o p oduce biomass composed o
mainly one mic obial species o ood applica ions. L. delb ueckii is
conside ed as a sa e o ganism, wi h p obio ic p ope ies, o ood and
eed applica ions due o i s longs anding use in he ood indus y as dai y
s a e (Aghababaie e al., 2011; Bei el e al., 2020) bu addi ionally i
ecei ed a QPS s a u e by EFSA and Gene ally Rega ded As Sa e (GRAS)
s a us by FDA (EFSA e al., 2022; FDA, 2012). In he case o ood in-
g edien s, supplemen a ion o his biomass could enhance la ou and
b ing immunos imulan p ope ies (Fang e al., 2020; Guglielmo i e al.,
2007). Fo li es ock eed, he biomass can enhance he eed con e sion
a es and ac as an an ibio ic eplacemen (Mo e al., 2022; Suda e al.,
2021).
Fig. 4. Rela i e abundance o he mic obial communi y composi ion o he op 13 abundan (≥2% ela i e abundance) ASVs assigned o he genus pe condi ion and
inoculum. Op ical pu i y o L-lac ic acid is ep esen ed by black do s.
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
7
3.3. Al e na i e lac ic acid p oduc ion scena ios can be achie ed h ough
selec ion o ope a ional pa ame e s
By ca e ully selec ing key ope a ional pa ame e s, we can maximize
all LA p oduc ion me ics. Speci ically, lowe ing he empe a u e om
55 ◦C o 45 ◦C enhanced LA p oduc i i y, concen a ion, selec i i y, and
yield al hough i led o a dec ease in op ical L-LA pu i y. Ope a ing a
ci cumneu al pH ange o 6.0–6.5, a 45 ◦C, signi ican ly imp o ed he
o e all LA pe o mance. Addi ionally, unning he eac o a low HRT
was he decisi e ac o o ob ain high LA selec i i y.
Ine i ably ade-o s exis be ween he LA p oduc ion pa ame e s
which should be conside ed when designing ope a ions o ce ain LA
applica ions. In he ollowing sec ion, we e alua e a numbe o ope a-
ional pa ame e s associa ed wi h di e en p oduc ion scena ios wi hin
he LA bio e ine y (e.g. bulk pla o m chemical, biopolyme p oduc ion
o in e media e o mic obial p o ein (MP)). The key ade-o be ween
p oduc i i y and yield is isualized in Fig. 5, which con ains ou esul s
as well as da a ound in li e a u e (Table S1). All da a ob ained om non-
axenic and axenic LA e men a ions ha e been o ganized by ope a ional
empe a u e (37–55 ◦C) and pH (5.0–6.5).
I he aim is o p oduce LA as a pla o m chemical, ope a ion should
a ge a combina ion o maximum LA yield and LA p oduc i i y ( op
igh co ne o Fig. 5) since his minimizes o e all p oduc ion cos s
(Do an, 2013; Po hakos e al., 2018). Howe e , a ade-o be ween bo h
LA me ics can be obse ed (Fig. 5). To ensu e high yields while main-
aining p oduc i i y a he highes possible le els, he p ocess should be
ope a ed a 45 ◦C, HRT 10h and a pH be ween 5.5 and 6.5 (g ey a ea in
Fig. 5). In ou es s, his led o a combina ion o high LA selec i i y
(96.2–97.8%), yield (85.9–100.9%), and concen a ion (46.6–51.1 g/L),
while achie ing ela i ely high LA p oduc i i y (5.0–5.5 g/L/h)
compa ed o o he s udies.
Fo LA applica ions equi ing high op ical pu i y (e.g. PLA and
pha maceu ical applica ions), pu e cul u es a e cu en ly employed o
p oduce ei he L-LA o D-LA. Howe e , ou s udy demons a ed ha
high op ical pu i y and LA yield can also be achie ed unde non-axenic
condi ions. Speci ically, ope a ing a empe a u es o 45–50 ◦C and pH
5.5 (blue squa e in Fig. 5) is op imal o D-LA p oduc ion wi h an op ical
pu i y o 85.9–91.3% and a LA yield o 87.4–100.9%. In con as ,
ope a ing a 45 ◦C and pH 6.5 esul s in L-LA p oduc ion a an op ical
pu i y o 95.6% and a LA yield o 93.5–99.0% (blue squa e in Fig. 5).
In case he p oduced LA is in ended o be used as a subs a e in
biological p ocesses (e.g. MP) maximizing ela i e abundance o LAB
should be p io i ized o e yield and op ical pu i y since he uncon e ed
subs a e will be ully consumed in he subsequen p ocess. High
abundance o LAB can p o ide an added ad an age by enhancing he
nu i ional alue o MP h ough i s p obio ic p ope ies. To emo e any
conce ns on he sa e y o he mic obial biomass p oduced unde non-
axenic condi ions, he ope a ional condi ions should be ca e ully
selec ed. In his espec , 45 ◦C a pH 6.0 and HRT 6h and 50 ◦C a pH 5.5
is ecommended o subsequen MP p oduc ion as i showed 90.3–98.6%
(b own squa e in Fig. 5) ela i e abundance o L. delb ueckii. This species
has a longs anding use in he dai y indus y and as a p obio ic which
could inc ease he unc ional p ope ies o he MP p oduc , i used as li e
cul u es.
3.4. A p elimina y assessmen o he po en ial o alo ise cheese whey
pe mea e in di e en lac ic acid-based applica ions
CWP has a signi ican po en ial as a eeds ock o p oduce biobased
p oduc s. A p elimina y assessmen was conduc ed o explo e he po-
en ial o his s eam o p oducing LA o LA-de i ed p oduc s (e.g., bulk
pla o m chemicals, PLA, o MP), ocusing on cu en p oduc ion capa-
bili ies and po en ial CO
2
educ ions wi hin he EU (see SI sec ion S4).
Based on a scena io using LA as a pla o m chemical and he esul s
Fig. 5. P oduc i i y and yield o lac ic acid (LA) p oduc ion ob ained om di e en s udies. The mixed communi ies shown achie ed a LA selec i i y o 95% o
abo e. Resul s a e o ganized based on eac o pH and empe a u e. As e isks indica e esul s ob ained in his s udy. The colou ed a eas show clus e s o da a poin s
ha achie ed simila alues o key pe o mance indica o s. Each di e en colou ed a ea is a isualisa ion o di e en p oduc ion scena ios: he g ey a ea shows da a
whe e LA could se e as pla o m chemical, he blue a ea shows da a whe e op ical pu e L-LA o D-LA can be achie ed, and he b own a ea a ange da a o LA
p oduc ion wi h a sequen ial p oduc ion s ep such as mic obial p o ein. (Fo in e p e a ion o he e e ences o colou in his igu e legend, he eade is e e ed o
he Web e sion o his a icle.)
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
8
ob ained in ou s udy, he cu en amoun o CWP being p oduced can
yield up o 2,600 k on o LA (see S4. Calcula ions). In he case LA is used
o PLA p oduc ion bo h LA enan iome s a e impo an as hey can yield
polyme s wi h di e en mechanical p ope ies. Taking in accoun he
yields and op ical pu i y ob ained in ou s udy, 1,400 k on poly-L-LA o
1,200 k on o poly-D-LA can be ob ained h ough he alo isa ion o
Eu ope’s CWP. To pu his in pe spec i e, annually he EU p oduces 65
k on o LA and 7 k on o PLA (Balla e al., 2021; Spek eijse e al., 2019).
This p oduc ion is based upon he u ilisa ion o ood-g ade suga s
de i ed om c ops which is in di ec compe i ion wi h ood p oduc ion
while causing en i onmen al damage. The u ilisa ion o CWP could
comple ely ci cum en he need o using ood-g ade eeds ocks, while
educing cos s, o p oduce plas ics o chemicals while inc easing he EU
annual p oduc ion 40 imes o LA as pla o m chemical and 125 imes
o PLA p oduc ion (see S4. Calcula ions). I p oduc ion o PLA on CWP
eplaced ossil uel-based PET, his could educe CO
2
emissions by 1.9
million on CO
2
equi alen s each yea (Mo e i e al., 2021). In addi ion,
he use o CWP as eeds ock can educe up o 16.5% o he acidi ica ion,
33.0% o he eu ophica ion, and up o 51.5% o he o al wa e use
ela ed o eeds ock p oduc ion o indus ial LA e men a ion compa ed
o LA e men a ions elying on ood-g ade eeds ocks (Mo e i e al.,
2021).
In he case CWP is used o p oduce MP, sa e y is o p ime impo ance
and only eac o condi ions we e conside ed whe e he communi y
consis ed ou o 98.0% L. delb ueckii. Unde hese condi ions 70% o he
lac ose was con e ed o LA. This scena io could lead o he p oduc ion
o 0.4 million on o p o eins con aining all essen ial amino acids,
assuming a heo e ical biomass yield o 0.63 gDM/g-C elec on dono
consumed (E dman e al., 1977; P´
aez e al., 2008). In case he MP is used
as eed, his could esul in he educ ion o p o ein impo in he EU by
2.23% (see S4. Calcula ions). This could educe up o 2.5 million on
CO
2
equi alen s annually, depending on he li es ock (Mancuso e al.,
2019; Sha ma e al., 2018; Tallen i e e al., 2018). Using he MP di ec ly
o p o ein eplacemen in ood p oduc s could esul in a 2.77%
educ ion o he annual EU p o ein consump ion. As a esul , his could
ha e an annual CO
2
o se o 18 million on CO
2
equi alen s (o 4.87% o
o al CO
2
equi alen s ela ed o animal p o ein p oduc ion) (see S4.
Calcula ions). In summa y, CWP can be used as a eeds ock o a lexible
LA pla o m and esul in high quan i ies o LA, PLA, and MP ha could
educe en i onmen impac caused h ough cu en p oduc ion o ood,
eed, chemicals, o plas ics om p ima y esou ces.
4. Conclusions
In his wo k, we showed ha non-axenic LA e men a ion can be
s ee ed h ough ca e ul selec ion o eac o pa ame e s (i.e. empe a-
u e, pH, and HRT). Dec easing he empe a u e om 55 ◦C o 45 ◦C,
imp o ed LA p oduc i i y, yield, and selec i i y. A pH o 6.5 (a 45 ◦C)
yielded he highes LA p oduc i i y (27.4 g/L/h) eco ded o da e (Y
LA
:
70.0%, C
LA
: 44.2 g/L, OP
L-LA
: 90.8%). Ci cumneu al pH (6.5) a ou ed
he o ma ion o L-LA, wi h a maximal op ical pu i y o 98.3% (P
LA
: 16.7
g/L/h, Y
LA
: 88.7%, C
LA
: 51.5 g/L), while a mildly acidic pH esul ed in a
maximal D-LA op ical pu i y o 93.2% (P
LA
: 7.3 g/L/h, Y
LA
: 87.4%, C
LA
:
46.0 g/L). The high D-LA op ical pu i y coincided wi h a 99.2% ela i e
abundance o a single Lac obacillus (ASV 1) species, e en in non-axenic
condi ions. Fu he mo e, high L-LA op ical pu i y was associa ed wi h
an inc eased ela i e abundance o h ee di e en S ep ococcus ASVs. In
addi ion o he esul s ob ained, we elabo a ed how ope a ional pa-
ame e s can be uned o mee di e en p oduc ion a ge s o applica-
ions. A p elimina y impac assessmen was p o ided o h ee LA
applica ions, as pla o m chemical, o PLA and MP p oduc ion).
CRediT au ho ship con ibu ion s a emen
B ech Delmoi i´
e: W i ing – o iginal d a , Visualiza ion, Me hod-
ology, In es iga ion, Fo mal analysis, Da a cu a ion, Concep ualiza ion.
My sini Saka ika: W i ing – e iew & edi ing, Supe ision, Me hodol-
ogy, Concep ualiza ion. Ko neel Rabaey: W i ing – e iew & edi ing,
Concep ualiza ion. Heleen De We e : W i ing – e iew & edi ing, Su-
pe ision, Concep ualiza ion. Albe e Reguei a: W i ing – e iew &
edi ing, Supe ision, Me hodology, Concep ualiza ion.
Decla a ion o compe ing in e es
The au ho s decla e ha hey ha e no known compe ing inancial
in e es s o pe sonal ela ionships ha could ha e appea ed o in luence
he wo k epo ed in his pape .
Acknowledgmen s
B.D. was suppo ed by he Resea ch Founda ion Flande s (FWO-
Vlaande en; G an numbe : 1S20323N). M.S. was suppo ed by Ghen
Uni e si y (BOF.PDO.2021.0036.01). A.R. acknowledges he suppo o
he Xun a de Galicia h ough a pos doc o al ellowship (ED481B-2021-
012). A.R. belongs o a Galician Compe i i e Resea ch G oup (GRC
ED431C 2021/37), co unded by ERDF (EU).
The au ho s g a e ully acknowledge Milcobel o p o iding he
subs a es used in his wo k.
Appendix A. Supplemen a y da a
Supplemen a y da a o his a icle can be ound online a h ps://doi.
o g/10.1016/j.jen man.2024.123529.
Da a a ailabili y
The aw as q iles ha se ed as a basis o he bac e ial communi y
analysis we e deposi ed in he Na ional Cen e o Bio echnology In-
o ma ion (NCBI) da abase (accession numbe : PRJNA949127). All
o he da a can be made a ailable upon eques .
Re e ences
Abdel-Rahman, M.A., Sonomo o, K., 2016. Oppo uni ies o o e come he cu en
limi a ions and challenges o e icien mic obial p oduc ion o op ically pu e lac ic
acid. J. Bio echnol. 236, 176–192. h ps://doi.o g/10.1016/j.jbio ec.2016.08.008.
Abdel-Rahman, M.A., Tashi o, Y., Sonomo o, K., 2013. Recen ad ances in lac ic acid
p oduc ion by mic obial e men a ion p ocesses. Bio echnology Ad ances 31 (6),
877–902. h ps://doi.o g/10.1016/j.bio echad .2013.04.002.
Abdel-Rahman, M.A., Tashi o, Y., Zendo, T., Sakai, K., Sonomo o, K., 2016. Highly
e icien L-lac ic acid p oduc ion om xylose in cell ecycle con inuous e men a ion
using En e ococcus mund ii QU 25. RSC Ad . 6 (21), 17659–17668. h ps://doi.o g/
10.1039/C5RA27579B.
Adambe g, K., Kask, S., Lah , T.-M., Paalme, T., 2003. The e ec o empe a u e and pH
on he g ow h o lac ic acid bac e ia: a pH-auxos a s udy. In . J. Food Mic obiol. 85
(1), 171–183. h ps://doi.o g/10.1016/S0168-1605(02)00537-8.
Aghababaie, M., Khanahmadi, M., Behesh i, M., Mi lohi, M., 2011. Tempe a u e and pH
Op imiza ion o he G ow h o Lac obacillus Delb ueckii Ssp Bulca icus on Whey
-based Medium.
Akao, S., Tsuno, H., Cheon, J., 2007. Semi-con inuous l-lac a e e men a ion o ga bage
wi hou s e ile condi ion and analysis o he mic obial s uc u e. Wa e Res. 41 (8),
1774–1780. h ps://doi.o g/10.1016/j.wa es.2007.01.013.
Ande sen, S.J., Hennebel, T., Gildemyn, S., Coma, M., Desloo e , J., Be on, J.,
Tsukamo o, J., S e ens, C., Rabaey, K., 2014. Elec oly ic memb ane ex ac ion
enables p oduc ion o ine chemicals om bio e ine y sides eams. En i on. Sci.
Technol. 48 (12), 7135–7142. h ps://doi.o g/10.1021/es500483w.
APHA, 2017. 2540B solids. In: S anda d Me hods o he Examina ion o Wa e and
Was ewa e . APHA P ess, pp. 1–23. h ps://doi.o g/10.2105/SMWW.2882.030.
A idan, O., G eene , T., 2021. P oduc ion o lac ic acid om o ganic was e using
composi ions o Bacillus coagulans spo es. h ps://pa en scope.wipo.in /sea ch/en/
de ail.js ?docId=WO2021191901.
Balla, E., Daniilidis, V., Ka lio i, G., Kalamas, T., S e anidou, M., Bikia is, N.D.,
Vlachopoulos, A., Koumen akou, I., Bikia is, D.N., 2021. Poly(lac ic acid): a e sa ile
biobased polyme o he u u e wi h mul i unc ional p ope ies— om monome
syn hesis, polyme iza ion echniques and molecula weigh inc ease o PLA
applica ions. Polyme s 13 (11), 1822. h ps://doi.o g/10.3390/polym13111822.
Ba man, S., Bha acha ya, S.S., Chand a Mandal, N., 2020. Chap e 3—Se a ia. In:
Ama esan, N., Sen hil Kuma , M., Annapu na, K., Kuma , K., Sanka ana ayanan, A.
(Eds.), Bene icial Mic obes in Ag o-Ecology. Academic P ess, pp. 27–36. h ps://doi.
o g/10.1016/B978-0-12-823414-3.00003-4.
B. Delmoi i´
e e al.
Jou nal o En i onmen al Managemen 373 (2025) 123529
9
