1
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
In i o and in i o e icacy o
combina ions o colis in and
di e en endolysins agains clinical
s ains o mul i-d ug esis an
pa hogens
Lucia Blasco1,9,11, An on Amb oa1,9,11, Rocio as oy1,9, Ines Ble io 1,9, Mi iam Moscoso1,10,
Lau a e nández-Ga cia1,9, Elena pe ez-nadales2,10, Felipe e nández-cuenca3,9,10,
Julian o e-cisne os2,10, Jesus o eo-iglesias4,9,10, An onio oli e 5,9,10, Ra ael can on6,9,10,
Tim Kidd7, Fe an na a o8,9, Elisenda Mi ó8,9, Al a o pascual3,9,10, Ge man Bou1,9,10,
Luis Ma ínez-Ma ínez2,9,10,11 & Ma ia omas1,9,10,11 ✉
The eme gence o mul id ug esis an (MDR) pa hogenic bac e ia is jeopa dizing he alue o
an imic obials, which had p e iously changed he cou se o medical science. In his s udy, we
iden i ied endolysins ElyA1 and ElyA2 (GH108-PG3 amily), p esen in he genome o bac e iophages
Ab1051Φ and Ab1052Φ, espec i ely. The mu aly ic ac i i y o hese endolysins agains MDR clinical
isola es (Acine obac e baumannii, Pseudomonas ae uginosa and Klebsiella pneumoniae) was es ed
using he u bidi y educ ion assay. Minimal inhibi o y concen a ions (MICs) o endolysin, colis in
and a combina ion o endolysin and colis in we e de e mined, and he an imic obial ac i i y o each
ea men was con i med by ime kill cu es. Endolysin ElyA1 displayed ac i i y agains all 25 s ains
o A. baumannii and P. ae uginosa es ed and agains 13 ou o 17 s ains o K. pneumoniae. Endolysin
ElyA2 did no display any such ac i i y. The combined an imic obial ac i i y o colis in and ElyA1
yielded a educ ion in he colis in MIC o all s ains s udied, excep K. pneumoniae. These esul s
we e con i med in i o in G. mellonella su i al assays and in mu ine skin and lung in ec ion models.
In conclusion, combining colis in (1/4 MIC) wi h he new endolysin ElyA1 (350 µg) enhanced he
bac e icidal ac i i y o colis in in bo h in i o and in i o s udies. This will po en ially enable educ ion
o he dose o colis in used in clinical p ac ice.
1Mic obiology Depa men -Resea ch Ins i u e Biomedical A Co uña (INIBIC), Hospi al A Co uña (CHUAC),
Uni e si y o A Co uña (UDC), A Co uña, Spain. 2Uni o Mic obiology, Uni e si y Hospi al Reina So ía, Depa men
o Mic obiology, Uni e si y o Có doba, Maimonides Biomedical Resea ch Ins i u e o Co doba (IMIBIC), Co doba,
Spain. 3Clinical Uni o In ec ious Diseases, Mic obiology and P e en i e Medicine, Hospi al Uni e si a io Vi gen
Maca ena / Depa men o Mic obiology and Medicine, Uni e si y o Se ille/ Biomedicine Ins i u e o Se ille (IBIS),
Se ille, Spain. 4Re e ence and Resea ch Labo a o y o An ibio ic Resis ance and Heal h Ca e In ec ions, Na ional
Cen e o Mic obiology, Ins i u e o Heal h Ca los III, Majadahonda, Mad id, Spain. 5Mic obiology Depa men -
Resea ch Ins i u e Biomedical Islas Balea es (IdISBa), Hospi al Son Espases, Palma de Mallo ca, Spain. 6Mic obiology
Depa men -Resea ch Ins i u e Biomedical Ramón and Cajal (IRYCIS), Hospi al Ramón and Cajal, Mad id, Spain.
7School o Chemis y and Molecula Biosciences and Child Heal h Resea ch Cen e, The Uni e si y o Queensland,
B isbane, QLD, Aus alia. 8Mic obiology Depa men -San Pau Hospi al, Ba celona, Spain. 9S udy G oup on
Mechanisms o Ac ion and Resis ance o An imic obials (GEMARA) o he Spanish Socie y o In ec ious Diseases and
Clinical Mic obiology (SEIMC), Mad id, Spain. 10Spanish Ne wo k o Resea ch in In ec ious Diseases (REIPI), Se ille,
Spain. 11These au ho s con ibu ed equally: Lucia Blasco, An on Amb oa, Luis Ma ínez-Ma ínez and Ma ia Tomas.
✉e-mail: MA.del.Ma .T[email p o ec ed]
open
2
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
The wo ldwide eme gence o mul id ug esis an (MDR) mic oo ganisms ha a e e ac o y o ea men wi h
cu en he apeu ic agen s has highligh ed he u gen need o new classes o an imic obial agen s1. The Wo ld
Heal h O ganiza ion (WHO) has ecen ly published a lis o “p io i y pa hogens” which includes hose mic o-
o ganisms ha a e conside ed a se ious h ea o human heal h and o which new an i-in ec i e ea men s a e
u gen ly needed. The lis includes ca bapenem- esis an A. baumannii, P. ae uginosa and K. pneumoniae clinical
isola es2.
One consequence o he eme gence o he MDR bac e ia is a e u n o he use o p e iously abandoned an imi-
c obials. This is he case wi h colis in (polymyxin E), a ca ionic pep ide which dis u bs he s abili y and inc eases
he pe meabili y o he ou e memb ane ia elec os a ic in e ac ions and ca ionic displacemen o he lipopoly-
saccha ide. Al hough colis in exe s an imic obial e ec s, i also has neph o oxic e ec s and has g adually been
abandoned and subs i u ed by o he , be e - ole a ed an ibio ics3,4. Combining new an imic obial agen s wi h old
an ibio ics such as colis in is a new s a egy in he de elopmen o no el ea men s agains MDR mic oo ganisms.
In ecen yea s, a no el d ug disco e y app oach has explo ed endolysin enzymes (also e e ed o as enzybi-
o ics), which a e encoded by bac e iophages ( i uses which in ec bac e ia) (5). Endolysins a e ac i ely p oduced
du ing he ly ic cycle and exe an ibac e ial ac i i y by deg ading pep idoglycan in he bac e ial cell wall5,6.
Endolysins a e highly e ol ed enzymes p oduced by bac e iophages o diges he bac e ial cell wall a he end
o hei eplica ion cycle and elease he phage p ogeny. Endolysins a ge he in eg i y o he cell wall and a ack
one o he majo bonds in he pep idoglycan laye . They can be classi ied in o i e g oups acco ding o he clea -
age si e: N-ace yl-β-D-mu amidase (lysozymes); N-ace yl-β-D-glucosaminidases (glycosidases); ly ic ansglyco-
sylase; N-ace ylmu amoyl-L-alanine amidases and L-alanoyl-D-glu ama e endopep idases7,8.
Endolysins a e good candida es as new an imic obial agen s agains G am-posi i e bac e ia, in which he
pep idoglycan laye o he cell wall is exposed o he medium. Se e al s udies ha e e alua ed he po en ial use o
endolysins agains G am-posi i e bac e ia such as S aphylococcus au eus, S ep ococcus agalac iae, S ep ococcus
pneumoniae and S ep ococcus pyogenes in animal models o human in ec ions and diseases9–16. In G am-nega i e
bac e ia, he ou e memb ane ac s as a ba ie o many endolysins, and e y ew endolysins wi h exogenous ac i -
i y agains G am-nega i e bac e ia ha e been desc ibed (many a e bio echnologically enginee ed)17–20. Endolysins
can a ack G am-nega i e bac e ia when he ou e memb ane is p e iously pe meabilized wi h agen s such as
EDTA, which des abilizes he lipopolysaccha ides o he ou e memb ane; howe e , he combina ion o endolysin
and EDTA is limi ed o opical ea men o localized in ec ions21,22. In he sea ch o al e na i e me hods o kill-
ing MDR bac e ia such as A. baumannii, P. ae uginosa and K. pneumoniae, a ious esea che s ha e conside ed
inc easing he mu aly ic ac i i y o endolysins by combining hem wi h di e en an ibio ics o ake ad an age o
syne gis ic esponses22,23.
In his s udy, we iden i ied and cha ac e ized an endolysin, named ElyA1, isola ed om he A. baumannii
Ab105 (ROC0034a) bac e iophage Ab1051Φ. The endolysin displayed mu aly ic ac i i y agains a b oad spec-
um o MDR o ganisms. In addi ion, combining endolysin ElyA1 wi h colis in (polymyxin E) enhanced he
suscep ibili y o he es ed s ains by a leas ou imes ( ela i e o he suscep ibili y o colis in alone), hus high-
ligh ing he po en ial o endolysin ElyA1 as a candida e an ibac e ial agen . This e ec was con i med by an in
i o es , in which he su i al o he G. mellonella la ae inc eased when colis in (¼ MIC) was supplemen ed
wi h endolysin ElyA1. Ano he endolysin om he same amily, named ElyA2, was iden i ied in he A. baumannii
Ab105 bac e iophage Ab1052Φ, bu did no display mu aly ic ac i i y.
Resul s
Iden i ica ion o endolysins ElyA1 and ElyA2. The 546 bp gene coding o endolysin ElyA1 was iden-
i ied as an ORF (Open Reading F ame) encoding a p o ein o 181 aa (GenBank: ALJ99090.1) and molecula
weigh , 20.22 kDa (Fig.1). The p o ein sequence was analysed wi h In e P oScan and classi ied as a lysozyme
(N-ace ylmu amidase) wi h a C- e minal domain co esponding o he glycoside hyd olase supe amily 108 and
also a pep idoglycan binding domain PG3 a he N- e minal end.
P o ein homology analysis e ealed a high le el o homology (>80%) wi h a g oup o 9 endolysins om A.
baumannii bac e iophages belonging o he same p o ein amily as ElyA120.
The 543 bp gene coding o endolysin ElyA2 was iden i ied as an ORF encoding a p o ein o 180 aa (GenBank:
ALJ99174.1) and molecula weigh 20.19 kDa (Fig.1). The sequence analysis e ealed ha he ElyA2 p o ein
is also a lysozyme (N-ace ylmu amidase), wi h a C- e minal domain co esponding o he glycoside hyd olase
supe amily 108, and also a pep idoglycan binding domain PG3 a he N- e minal end.
Like he ElyA1 p o ein, his enzyme displays a high deg ee o homology (>80%) wi h he same g oup o 9
endolysins and also 90% homology wi h he ElyA1 p o ein20.
Cha ac e iza ion o endolysin mu aly ic ac i i y. In he ini ial sc eening o he mu aly ic ac i i y o
he pu i ied endolysin ElyA1 in he o e lay pla es wi h G am-nega i e bac e ia, a halo appea ed a ound he lysis
zones o bo h s ains o A. baumannii es ed (Fig.2a).
The mu aly ic ac i i y o his enzyme was cha ac e ized using he G am-nega i e bac e ia A. baumannii
Ab105 as subs a e, as his is he hos s ain o he phage Ab105Φ1. The enzyma ic ac i i y was measu ed a e
incuba ion a di e en empe a u es and pH. The maximum ac i i y was ob ained a e incuba ion o 10 min a
pH8.5 and 37 °C (Fig.2b,c). In addi ion, he mu aly ic ac i i y on he Ab105 cells was assayed di ec ly o a e
ea men o he cells wi h EDTA o pe meabilize he ou e memb ane. Howe e , no ac i i y was de ec ed when
he enzyme was added di ec ly o he cells whose ou e memb ane had no been pe meabilized wi h EDTA, and
in his case he cells also ended o agg ega e (da a no shown).
The an ibac e ial assays showed a b oad ly ic spec um o ac i i y agains he s ains o he h ee species es ed
(Fig.3). As expec ed because o he o igin o he A. baumannii endolysin, he ac i i y was highes among he 25 A.
baumannii s ains es ed. Al hough he ac i i y was mo e a iable in P. ae uginosa, mu aly ic ac i i y agains all o
3
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
he s ains was de ec ed. Finally, endolysin ElyA1 was ac i e agains 13 o he 17 K. pneumoniae s ains, al hough
a lowe le els han in A. baumannii and P. ae uginosa. The s ains o he h ee species es ed belonged o di e en
s ain ypes (STs), bu he suscep ibili y o endolysin ElyA1 was no co ela ed wi h he ST.
Tes s o he mu aly ic ac i i y o endolysin ElyA2 did no de ec any ac i i y unde any he condi ions assayed.
On he con a y, his enzyme induced agg ega ion o he cells a all he enzyme concen a ions es ed, bo h in he
cells ea ed p e iously wi h EDTA and in hose wi h an in ac ou e memb ane.
Combined ac i i y o endolysin ElyA1 and colis in in in i o assays. As ElyA1 is only ac i e when
he ou e memb ane o he a ge bac e ial cell is solubilized, he MIC o he endolysin could no be de e mined
using he mic odilu ion checke boa d es . We he e o e aimed o de ec any dec ease in he colis in MICs when
used in combina ion wi h endolysin ElyA1. The addi ion o endolysin ElyA1 yielded a ou old educ ion in he
colis in MICs in ou o he six s ains es ed (A. baumannii GMA001 and PON001, P. ae uginosa AUS531 and
K. pneumoniae KP17) (Fig.4). By con as , only a wo old educ ion in he colis in MIC was obse ed wi h P.
ae uginosa AUS601 and no dec ease wi h K. pneumoniae KP16. The la e was consis en wi h he lack o enzy-
ma ic ac i i y obse ed in he an ibac e ial assays (Fig.4). Finally, no an imic obial ac i i y was de ec ed when
he combina ion was es ed in he colis in esis an isola es (da a no shown).
Figu e 1. Genome o he bac e iophages Ab105Φ1 (GenBank: KT588074.1) and Ab105Φ2 (GenBank:
KT588075.2) by igu e modi ied wi h PHAST so wa e (h p://phas .wisha lab.com) (60). SDS-PAGE
pu i ica ion o he endolysins ElyA1 and ElyA2 (ch oma og aphic s udy).
4
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
The esul s o he ime kill cu e assay con i med he esul s o he mic odilu ion checke boa d es (Fig.4).
A 2 log educ ion in g ow h o bo h o he A. baumannii s ains and P. ae uginosa AUS531 a e 6 hou s in he
cul u e wi h 1/4 o colis in and endolysin ElyA1 was obse ed, indica ing a syne ge ic eac ion be ween colis in
and endolysin ElyA1. By con as , he e was no educ ion in g ow h in he K. pneumoniae KP17 cul u e.
Ac i i y o endolysin/colis in combina ions in in i o assays:
(a) Mo ali y in he in i o Galle ia mellonella model (Fig.5a)
La ae o he wax mo h we e in ec ed wi h clinical s ain A. baumannii GMA001. Su i al o in ec ed
la ae ea ed wi h colis in (¼MIC) in combina ion wi h he ElyA1 (25 µg/ml) was signi ican ly highe
(p < 0.05) ha ha o la ae ea ed wi h colis in only (¼ MIC). T ea men wi h he combina ion o colis in
(¼ MIC) and ElyA2 (25 µg/ml) did no yield signi ican di e ences (p > 0.05) ela i e o ea men wi h he
colis in ea men , as ElyA2 did no display mu aly ic ac i i y.
(b) E icacy o ElyA 1 in he mu ine skin model (Fig.5b)
Mice supe icial skin wounds we e in ec ed wice (on wo consecu i e days) wi h clinical s ain A. bauman-
nii GMA001. The wounds we e ea ed wi h colis in in combina ion wi h di e en doses o ElyA1 (50 µg
and 350 µg), a colis in con ol o a bu e con ol. The e ec i eness o he ea men s was es ablished by
coun ing he o al numbe o CFUs in he skin wound. The cell coun s we e signi ican ly lowe (p ≤ 0.05;
S uden ’s - es ) in he colis in combina ion ea men s (wi h bo h doses o ElyA1) han in he bu e
con ol. The cell coun s in he 350 µg ElyA1 plus colis in ea men we e also signi ican ly lowe (p ≤ 0.05;
S uden ’s - es ) han in he colis in con ol.
(c) E icacy o ElyA1 in ea men o lung in ec ion (Fig.5c)
In ec ed mouse lungs we e only ea ed wi h he combina ion o colis in and 350 µg ElyA1, as colis in plus he
lowe dose o endolysin (50 µg) did no display any ac i i y in he skin in ec ion model.
Lung CFU coun s we e signi ican ly lowe (p ≤ 0.05; S uden ’s - es ) in he mice ea ed wi h he combina ion
o colis in and ElyA1 han in he mice ea ed wi h bu e o wi h colis in. The e we e no signi ican di e ences in
he CFU coun s be ween he bu e con ol g oup and he colis in con ol g oup.
Discussion
The disco e y and de elopmen o no el an imic obial agen s o ea in ec ions caused by he “p io i y” g oup o
pa hogens is a challenge acing he medical and esea ch communi y2.
Enzybio ics ha e become he ocus o a en ion o many esea ch g oups wo ldwide. Endolysins (one ype o
enzybio ics) a e species o genus-speci ic enzymes ha ac by hyd olysing he pep idoglycan laye o he bac e ial
cell wall. The e a e no epo s o bac e ia de eloping esis ance o endolysins, which is a p oblem in bo h an ibi-
o ic he apy and phage he apy16. Mo eo e , endolysins ha e been ecognized in he US “Na ional Ac ion Plan o
Comba ing An ibio ic- esis an Bac e ia”24, which iden i ied he use o “phage-de i ed lysins o kill speci ic bac e-
ia while p ese ing he mic obio a” as a key s a egy o educe he de elopmen o an imic obial esis ance due o
Figu e 2. Cha ac e iza ion o enzyma ic ac i i y: (a) Mu aly ic ac i i y o ElyA1 was de e mined by spo ing
ElyA1 and endolysin bu e as a nega i e con ol in an o e lay o wo G am-nega i e Acine obac e baumannii
clinical isola es, MAR001 and PAU002; (b) pH ange and (c) empe a u e ange we e de e mined by he speci ic
ac i i y, measu ed as he di e ence in op ical densi y o he cul u e pe µg o enzyme and minu e.
5
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
he absence o oxici y in human cells25,26. Mo eo e , some endolysins ha e been ound o display ac i i y agains
sub-popula ions o mic obes27,28 ex ac ed om bio ilm29–31 and o be use ul in o he inno a i e ea men s.
The ou e memb ane o G am-nega i e bac e ia ac s as a ba ie p e en ing access o many endolysins o hei
na u al a ge , he pep idoglycan laye . Di e en s a egies ha e been used o add ess his p oblem, including sol-
ubiliza ion o he ou e memb ane wi h EDTA, modi ica ion o he endolysin PGs by dele ion o subs i u ion, and
he de elopmen o usion p o eins such as A ilysin-175 (A -175). This p o ein is made by using he endolysin
wi h a pep ide, success ully enabling he enzyme o pass h ough he ou e memb ane18,32,33. A -175 cons i u ed
by using an imic obial pep ide (AMP) sheep myeloid 29-amino acid pep ide (SMAP-29) wi h endolysin KZ144
displayed mu aly ic ac i i y in a P. ae uginosa isola e, and con inuous exposu e o A -175 did no lead o he
de elopmen o esis ance18. By i sel , SMAP-29 is cy o oxic o mammalian cells34; howe e , A -175 exhibi ed
li le oxici y in L-292 mouse connec i e issue.
Figu e 3. Speci ic ac i i y o endolysin ElyA1 es ed in clinical isola es om di e en mu lilocus sequence
ypes (STs) o h ee G am-nega i e membe s o he ESKAPE g oup: Acine obac e baumannii, Pseudomonas
ae uginosa and Klebsiella pneumoniae.
6
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
As a new s a egy, we combined he memb ane-des abilizing e ec o colis in, a ca ionic pep ide used as an
ac i e ou e memb ane agen (bu only as a “las -line” ea men due o conce ns abou i s neph o oxici y and
neu o oxici y35), and wo endolysins iden i ied by ou esea ch g oup and belonging o a lysozyme-like amily
(GH108-PG3) ne e be o e used as an imic obial ea men .
Figu e 4. In i o bac e icidal ac i i y o colis in in combina ion wi h endolysin ElyA1 measu ed by MIC and
ime kill cu es in Acine obac e baumannii s ains GMA001 (a) and PON001 (b); Pseudomonas ae uginosa
s ains AUS531 (c) and AUS601 (d); Klebsiella pneumoniae s ains KP17 (e) and KP16 ( ). The ime kill cu es
we e only cons uc ed o s ains in which he e was a ou old educ ion in colis in MICs ( ed squa e) when
used in combina ion wi h endolysin ElyA1 (yellow squa e).
7
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
In his s udy, we iden i ied wo endolysins, ElyA1 and ElyA2, ob ained om A. baumannii bac e iophage
Ab1051Φ and Ab105Φ2, a ailable in a collec ion o clinical s ains o A. baumannii isola ed du ing he II Spanish
Mul icen e S udy GEIH/REIPI-A.baumannii 2000–2010 (Accession numbe PRJNA422585, Genbank Umb ella
Biop ojec )36,37.
Endolysins ElyA1 and ElyA2 a e lysozyme-like p o eins wi h a ca aly ic domain and a cell wall binding
domain (CBD), esponsible o ecogni ion o he cell su ace ligands and a ini y o he bac e ial subs a e6,38.
This s uc u e is mos commonly ound in endolysins om bac e iophages ha a ge G am-posi i e bac e ia.
Howe e , he PG_3 domain p esen in endolysins ElyA1 and ElyA2 has been iden i ied in some G am-nega i e
bac e ia and in a g oup o nine endolysins isola ed om bac e iophages o A. baumannii; he domain shows high
homology wi h ElyA1 and belongs o he same amily (Fig.1)20,36,39. The p esen indings ega ding he molecula
cha ac e is ics and compa a i e genomes in bac e iophage endolysins con i m p e iously epo ed indings40.
The bac e iophages om which hese endolysins we e isola ed, Ab1051Φ and Ab105Φ2, occu in a la ge num-
be o clinical isola es o A. baummanii36. The cell wall binding domain has been shown o be esponsible o he
speci ici y and a ini y o he endolysins o i s subs a e39. Howe e , endolysin ElyA1 displayed a b oade spec-
um o ac i i y agains s ains o A. baumannii and many s ains o P. ae uginosa belonging o he same o de
(Pseudomonadales), and o lesse ex en agains some s ains o K. pneumoniae om ano he gammap o eobac-
e ial o de , En e obac e ales. In his case, he a ge o endolysin ElyA1, iden i ied in pep idoglycan (PG) binding
domains as a D-Asn40, is p obably conse ed among he Pseudomonadales, hus explaining he b oad spec um o
ac ion o his enzyme. In e es ingly, we we e no able o de ec mu aly ic ac i i y in endolysin ElyA2, because his
enzyme induces agg ega ion o he cells in i o. An agg ega i e e ec was p e iously desc ibed in he endolysin
phi12 isola ed om a S. au eus bac e iophage, al hough he cause o he e ec was unknown41. Au oagg ega ion
has been sugges ed o occu in en i onmen al s ess caused by oxins, an ibio ics, p eda ion o low nu ien s42.
In he p esen s udy, we used he ca ionic polymyxin an ibio ic colis in o o e come he impene abili y o
he ou e memb ane o endolysin ElyA1. Colis in dis u bs he ou e memb ane ia an elec os a ic in e ac ion
wi h lipopolysaccha ides and phospholipids p esen in he ou e memb ane4. The syne gis ic e ec o colis in
and endolysin LysABP-01 (a lysozyme-like p o ein om he GH19 amily) on A. baumannii has p e iously been
desc ibed22. Al hough he endolysin ElyA1 does no display exogenous ac i i y, because o i s inabili y o c oss
he ou e memb ane, his p oblem was la gely o e come when he enzyme was used in combina ion wi h colis in.
Figu e 5. In i o bac e icidal ac i i y o colis in in combina ion wi h endolysins ElyA1 and ElyA2. (a) Su i al
cu es o G. mellonella la ae in ec ed wi h A. baumannii clinical s ain GMA001 and ea ed wi h colis in
(1/4 MIC) and wi h colis in (1/4 MIC) combined wi h endolysin ElyA1 (25 μg/ml). Su i al cu es o G.
mellonella la ae in ec ed wi h A. baumannii clinical s ain GMA001 and ea ed wi h colis in (1/4 MIC) o
wi h colis in (1/4 MIC) combined wi h endolysin ElyA2 (25 μg/ml). This expe imen was ca ied ou wi h
an app op ia e su i al con ol. *S a is ically signi ican di e ences (p < 0.05) we e de e mined by G aham-
B eslow-Wilcoxon es (G aphPad P ism .6); (b) An imic obial ac i i y o endolysin ElyA1 in a mu ine skin
model. CFU quan i ica ion in homogenized mouse skin a e in ec ion wi h A. baumannii GMA001 and
ea men wi h colis in (1/4 MIC) in combina ion wi h di e en doses o endolysin ElyA1 (50 µg and 350 µg)
o wi h bu e o colis in (con ols). (c) An imic obial ac i i y o ElyA1 in a mu ine lung in ec ion model. CFU
quan i ica ion in lungs a e in ec ion wi h A. baumannii GMA001 and ea men pos -in ec ion wi h colis in in
combina ion wi h350 µg o ElyA1. * S a is ically signi ican di e ences (p < 0.05) we e de e mined by -S uden
es (G aphPad P ism .6).
8
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
The an imic obial ac i i y o he combined he apy was highe han o bo h subs ances used alone, o all o he
s ains es ed, excep he K. pneumoniae s ains. A educ ion in he colis in MIC o a leas ou old was obse ed
o all o he A. baumannii s ains es ed and o P. ae uginosa s ain AUS531, and a co esponding wo old
educ ion was obse ed o P. ae uginosa s ain AUS601. A educ ion in he colis in MIC was also ob ained o
K. pneumoniae s ain KP17, he s ain mos suscep ible o endolysin ElyA1. The inc eased an imic obial ac i i y
wi h endolysin ElyA1 and colis in was con i med wi h an almos 3 log educ ion in g ow h a e 6 h in all s ains
es ed, excep K. pneumoniae KP17. G ow h o he cul u e eached he same le el as in he con ol a e 24 h,
p obably due o deg ada ion o he enzyme and colis in, as p e iously epo ed22. In all o he s ains es ed, he
educ ion in he colis in MIC was consis en wi h he mu aly ic ac i i y o endolysin ElyA1 obse ed wi h hose
s ains. No an imic obial ac i i y was obse ed when his assay was conduc ed wi h colis in- esis an s ains,
p obably because o he inabili y o he enzyme o access he pep idoglycan laye , as he necessa y des abiliza ion
o he ou e memb ane by he colis in was no p oduced in hese isola es. Howe e , se e al mechanisms o esis -
ance o colis in ha e been desc ibed. In some mechanisms, he lipopolysaccha ide is modi ied o is no p oduced,
p e en ing binding o he colis in o he ou e memb ane. O he mechanisms include e lux pumps desc ibed in
A. baumannii and inhibi ion o espi a o y enzymes such as NADH oxidase in G am-posi i e bac e ia such as
Bacillus spp. and NADH quinone oxido educ ase in E. coli. The ac i i y o he enzyme is likely o be highe in he
bac e ia wi h colis in esis ance mechanisms di e en om hose in ol ing modi ica ion o he lipopolysaccha-
ides43–49. In Eu ope, he incidence o colis in esis an A. baumannii in in ensi e ca e uni s eached o e 23% due
o di e en mechanisms o esis ance such as al e a ions in he lipopolysaccha ide (LPS) as well as acquis ion o
mc genes50. Because o he possible inabili y o hese combina ions o inhibi colis in esis an s ains, u he
s udies mus be conduc ed wi h a ange o di e en bac e ia wi h di e en mechanisms o esis ance o colis in
wi h he aim o educing he colis in MIC in combina ion wi h endolysin ElyA1.
The esul s ob ained in i o we e con i med wi h hose o in i o assays, as he su i al o he in ec ed G. mel-
lonella la ae was highe when he wo ms we e ea ed wi h a combina ion o a educed ( ou old) MIC o colis in
and endolysin ElyA1 han wi h colis in alone. As a con ol, he same assay was pe o med wi h endolysin ElyA2,
in which no mu aly ic ac i i y was de ec ed, and he e we e no di e ences ela i e o ea men wi h colis in. As
in G. mellonella, he an imic obial ac i i y o ElyA1 was con i med in i o. A combina ion o colis in and 350 µg
o ElyA1 was used o ea he skin in ec ion and lung in ec ion in mice, yielding a signi ican educ ion in he
numbe o bac e ia ela i e o ea men wi h colis in alone.
In conclusion, his is he i s in i o and in i o s udy in which colis in has been combined wi h endolysin
ElyA1 (glycosyde hyd olase supe amily 108) o ea in ec ions caused by clinical MDR pa hogens. This ype
o ea men may enable a educ ion in he concen a ion o colis in used in an imic obial ea men s, hus also
educing he oxic side e ec s o he an ibio ic. The b oad spec um o ac ion o endolysin ElyA1 would enable he
inclusion o mo e MDR G am-nega i e bac e ia as a ge s o he combined an imic obial ea men .
Ma e ials and Me hods
S ains and cul u e condi ions. The bac e ial s ains and plasmids used in his s udy included 25 A. bau-
mannii MDR s ains belonging o 22 di e en sequence ypes (STs) (Table1). The s ains we e isola ed om colo-
nized o in ec ed pa ien s wi hin he amewo k o he II Spanish Mul icen e S udy, in which 45 Spanish hospi als
pa icipa ed (GEIH-REIPI Acine obac e baumannii 2000–2010, Genbank Umb ella Biop ojec accession numbe
PRJNA422585)36,37. The s ains included 25 MDR clinical s ains o P. ae uginosa (many included in CC274), all
o which we e isola ed om cys ic ib osis pa ien s, and 17 ca bapenemase-p oducing s ains o K. pneumoniae,
which we e isola ed in 20 Spanish hospi als du ing he EuSCAPE p ojec 51,52. Mo eo e , Esche ichia coli DH5α
and Rose a s ains we e used in cloning assays (Table1).
All s ains we e cul u ed in LB (Lu ia-Be ani) b o h a 180 pm and 37 °C. Fo solid medium, 2% o aga was
added o LB b o h. In he ans o ma ion assays, he medium was supplemen ed wi h 50 µg/ml o ampicillin.
S ain, Plasmid, P ime ,
S ain Desc ip ion, Cha ac e is ics and Sequence O igin and
Re e ence
S ain
Acine obac e baumannii 25 clinical isola es (22 STs) om he II Spanish Mul icen e S udy (GEIH-REIPI Acine obac e
baumannii 2000–2010) (Accession numbe Genbank PRJNA422585) 27
Pseudomonas ae uginosa 25 clinical isola es (ST274 [n = 15]; ST1089 [n = 3]; ST no known [n = 7]) 28
Klebsiella pneumoniae 17 clinical isola es belonging o 16 di e en STs 29
Esche ichia coli DH5αS ain using o cloning No agen
Esche ichia coli Rose a
pLys-S S ain o p o ein exp ession No agen
Plasmid
pET-28a Km ,, T7lac, His-Tag, T7-Tag, h ombine p o ease si e No agen
P ime s
Fo wa d 5′-AGTTCTGTTCCAGGGGCCCCATATGAACATTGAACAATATCTTGATGAA-3 This s udy
Re e se 5′-AGTGGTGGTGGTGGTGGTGCTCGAGTCACATTGATACTCGATTAGCAAT-3′This s udy
Table 1. Desc ip ion o he bac e ial s ains, plasmids and p ime s used in his s udy. Abb e ia ions: ST;
mul ilocus sequence ype.
9
Scien i ic RepoR S | (2020) 10:7163 | h ps://doi.o g/10.1038/s41598-020-64145-7
www.na u e.com/scien i ic epo s
www.na u e.com/scien i ic epo s/
Iden i ica ion and pu i ica ion o endolysins ElyA1 and ElyA2. Endolysin gene p edic ion, om
he genome o he bac e iophage Ab105Φ1 (GenBank: KT588074.1) and Ab105Φ2 (GenBank: KT588075.2)36,53
(Fig.1), was pe o med wi h he bioin o ma ic ools PHASTER (Phage Sea ch Tool Enhanced Release) and
RAST (Rapid Anno a ion Using Subsys em). P o ein homology analysis was pe o med by BLAST (Basic Local
Alignmen Sea ch Tool), Clus al Omega and MView. P o ein amilies we e assigned using In e P oScan, and he
domain g aphic was assigned wi h PROSITE MyDomains.
The endolysin genes we e ampli ied by PCR om he genomic DNA o A. baumannii Ab105 (which con ains
he DNA o he p ophages Ab105Φ1 and Ab105Φ2) and cloned in o he exp ession ec o pET-28a (No agen).
The ecombinan plasmids we e ans o med in o compe en E. coli DH5α cells (No agen) o DNA p oduc ion
and pu i ica ion, and he in eg i y o bo h cons uc s was e i ied by sequencing. All o he p ime s used a e lis ed
in Table1. Finally, he plasmids we e ans o med in o Esche ichia coli Rose a pLys-S cells (No agen) o exp ess
he p o ein.
A e induc ion wi h 1 mM IPTG, he cul u e (1 l) was g own a 30 °C o 5 h. The bac e ial cells we e eco e ed
by cen i uga ion (in a JLA 81000 o o , Beckman-Coul e , a 6 K pm o 15 min) and dis up ed by sonica ion
(Vib aCell 75042 sonica o , Bioblock Scien i ic, ip model CV33). The sample was cen i uged in a JA 25–50 o o
(Beckman-Coul e ), a 20 K pm o 30 min. The supe na an was il e ed using 0.45 µm sy inge-d i en il e s (Je
Bio il) and loaded in a His-T ap column (GE Heal hca e) equilib a ed wi h 350 mM NaCl, 50 mM T is pH 7.5,
1 mM TCEP and 10 mM Imidazole. The p o eins we e elu ed wi h 350 mM NaCl, 50 mM T is pH 7.5, 1 mM TCEP
and 150 mM Imidazole. A e concen a ion in an Amicon Ul acel 10,000 MCWO concen a o (Millipo e), he
sample was loaded in o a Supe dex 75 16/60 column (GE Heal hca e), equilib a ed wi h 150 mM NaCl, 20 mM
T is pH 7.5 and 1 mM TCEP. The p o ein was elu ed in a single peak. Finally, he pooled peak ac ions we e con-
cen a ed o 40 mg/ml, as p e iously desc ibed. The pu i ica ion p ocess was ca ied ou a 4 °C, and he pu i y
was de e mined by SDS-PAGE (Fig.1).
De e mina ion o he mu aly ic ac i i y o endolysins ElyA1 and ElyA2. Mu aly ic ac i i y was
de e mined using he G am-nega i e o e lay me hod desc ibed by Schmi z e al.54. B ie ly, wo clinical isola es
o A. baumannii, MAR001 and PAU002, we e g own o s a iona y phase (109 CFU/ml) in LB, pelle ed and esus-
pended in PBS bu e pH 7.4. Aga was added di ec ly o he bac e ial suspension a a concen a ion o 0.8%, and
he mix u e was au ocla ed o 15 min a 120 °C. The medium con aining he diso ganized cells and he exposed
pep idoglycan was solidi ied in Pe i dishes, and aliquo s (50 μg) o endolysin o he endolysin bu e (as a nega-
i e con ol) we e spo ed on he su ace.
The mu aly ic ac i i y was measu ed using as a ge a cul u e o A. baumannii Ab105 ea ed wi h EDTA in
o de o pe meabilize he ou e memb ane. An o e nigh cul u e o A. baumannii Ab105 was dilu ed 1:100 in LB
medium and g own o exponen ial phase (0.3–0.4 OD600nm). The cul u e was cen i uged (3000 g, 10 min), and
he esul ing pelle was esuspended in 20 mM T is-HCl bu e a pH 8.5 wi h 0.5 mM EDTA be o e being incu-
ba ed o 30 min a oom empe a u e. The pelle was eco e ed by cen i uga ion and washed wice in T is-HCl
bu e pH 8.5. Finally, he cells we e esuspended in 20 mM T is-HCl 150 mM NaCl pH8.5 and 25 µg/ml o endo-
lysin ElyA1. The ac i i y was measu ed by he u bidi y educ ion assay, as a dec ease in he op ical densi y meas-
u ed a a wa eleng h o 600 nm (OD600) a e incuba ion wi h shaking a 37 °C17. The OD600 was measu ed a
in e als o 5 minu es o a pe iod o 20 minu es and he ime poin o he highes ac i i y was es ablished. The
op imal pH and empe a u e o endolysin ac i i y we e de e mined in he u bidi y educ ion assay. The eac ion
was ca ied ou as p e iously desc ibed, wi h he T is-HCl a di e en pH ( ange 6.5 o 9) and empe a u e ( oom
empe a u e, 30 °C and 37 °C).
An ibac e ial assays. The an ibac e ial ac i i y o he endolysin was assayed wi h all o he 67 clinical s ains
o A. baumannii, P. ae uginosa and K. pneumoniae (Table1). The ac i i y was de e mined using he u bidi y
educ ion assay, as p e iously desc ibed, a pH 8.5 and 37 °C. The incuba ion imes in he p esence o EDTA a -
ied acco ding o he species assayed: 30 min o A. baumannii and K. pneumoniae and 15 min o P. pneumoniae.
B o h mic odilu ion checke boa d assay and mic odilu ion es o de e mine minimum inhib-
i o y concen a ions (MICs). This assay was conduc ed wi h he s ains disaplying he highes and he
lowes suscep ibili y o endolysin. All he s ains es ed we e suscep ible o colis in(Table S1), excep h ee s ains,
which we e colis in esis an : A. baumannii SOF004b, P. ae uginosa AUS034 and K. pneumoniae KP2. The e ec
o he in e ac ion be ween endolysin and colis in was de e mined by he mic odilu ion checke boa d assay. Se en
se ial double dilu ions o endolysin and 6 o colis in we e made wi h Muelle -Hin on B o h (MHB) in he wells o
a 96-well mic o i e pla e. The wells we e hen inocula ed wi h he es cul u e o a inal concen a ion o 105 col-
ony o ming uni s (c u/ml). The MICs o colis in (0 o 2 µgml−1) and o he ElyA1 p o ein (3.125 o 200 µgml−1)
we e assayed independen ly in he same pla e. The MIC was de e mined as he concen a ion o an imic obial
agen in he well in which no isible g ow h o bac e ia was obse ed a e incuba ion o 24 h a 35 °C.
Time kill cu e assay. Time kill cu e assays we e ca ied ou wi h hose s ains in which he colis in MIC
in he colis in-ElyA1 combina ions was dec eased by a leas ou old in he checke boa d assays. The assay was
conduc ed acco ding o p e iously desc ibed echniques55. Flasks o LB con aining colis in and colis in plus endo-
lysin a he concen a ion indica ed in he checke boa d assay we e inocula ed wi h a 1:100 dilu ion o an o e -
nigh cul u e in s a iona y phase o he es ed s ain and incuba ed a 37 °C and 180 pm in a shaking incuba o .
Aliquo s we e emo ed a e 0, 6 and 24 h and we e se ially dilu ed and pla ed o p oduce colony o ming uni s
(c u). Syne gy was es ablished when a ≤ 2 −log10 dec ease in cells coun s a 6 o 24 h in he an imic obial combi-
na ion ela i e o he mos ac i e single agen was obse ed. No e ec was conside ed o ha e occu ed when he
coun s we e <2−log 10 lowe o highe ela i e o he cul u e wi h he single agen . An agonism was de ined when
