The Jou nal o Clinical In es iga ion
RESEARCH ARTICLE
5474 jci.o g Volume 129 Numbe 12 Decembe 2019
In oduc ion
Pe sis en immune ac i a ion and in lamma ion a e hallma ks o
ch onic HIV/SIV in ec ion and s ong p edic o s o disease p o-
g ession, independen o plasma i al loads (pVLs), o pe iphe al
CD4+ T cell coun s (1). They also pe sis in indi iduals wi h HIV/
SIV and nonhuman p ima es (NHPs) on an i e o i al he apy
(ART), p edic ing poo immune es o a ion, accele a ed aging, and
de elopmen o non–AIDS-associa ed como bidi ies (2–5). The
mechanisms o HIV/SIV-associa ed immune ac i a ion and in lam-
ma ion a e he subjec o in ense esea ch, wi h key de e minan s
including pe sis en i us eplica ion, coin ec ions, dys egula ion
o T cell homeos asis, and mic obial ansloca ion (4, 6–8). Mean-
while, o he unknown ac o s likely con ibu e o HIV/SIV-associ-
a ed ch onic immune ac i a ion and in lamma ion. Among hese,
die a y habi s may a ec in lamma ion h ough mul iple pa hways
(9), al e ing he pa hogenesis o ch onic HIV/SIV in ec ion. Li le is
known, howe e , wi h ega d o hese e ec s.
Consuming a high- a die (HFD) is a leading isk ac o o
obesi y and ype 2 diabe es (10), which a e bo h cha ac e ized by
ch onic in lamma ion (10), ma ked by ele a ed TNF-α, IL-1, IL-6,
and accumula ion o p oin lamma o y mac ophages in he adipose
issue (10). A HFD al e s he gas oin es inal (GI) mucosa, dis up s
gu mic obio a, and induces in es inal damage. These condi ions
all lead o mic obial ansloca ion (11–13), u he boos ing ch onic
immune ac i a ion and in lamma ion, hus accele a ing HIV/SIV
disease p og ession.
In humans, Wes e n die s a e associa ed wi h an inc eased isk
o nonalcoholic a y li e disease (NAFLD) (14). The impac o
Wes e n die on li e heal h is disce nible wi hin weeks (15). In he
Uni ed S a es, NAFLD became a signi ican heal h p oblem in male
adolescen s who equen ly consume Wes e n die s (15). HFD can
also induce NAFLD in mice and a s, along wi h signi ican ele a-
ion in TNF-α, NF-κB, and oxida i e s ess. These li e al e a ions
co ela e wi h hickening o he ca o id in ima (16), sugges ing a
possible connec ion be ween li e pa hology and ca dio ascula
(CV) disease. In humans, h ough ele a ed in lamma ion ia he
TLR4/NF-κB signaling pa hway, HFD causes enal, e inal, and
b ain unc ion impai men s (17). As li e , enal, and neu ological
diso de s a e also common HIV-associa ed como bidi ies, HFD
may inc ease he isk o o he se e i y o HIV como bidi ies.
Consuming a high- a die (HFD) is a isk ac o o obesi y and diabe es; bo h o hese diseases a e also associa ed
wi h sys emic in lamma ion, simila o HIV in ec ion. A HFD induces in es inal dysbiosis and impai s li e unc ion and
coagula ion, wi h a po en ial nega i e impac on HIV/SIV pa hogenesis. We adminis e ed a HFD ich in sa u a ed a s and
choles e ol o nonpa hogenic (A ican g een monkeys) and pa hogenic (pig ailed macaques) SIV hos s. The HFD had a
nega i e impac on SIV disease p og ession in bo h species. Thus, inc eased cell-associa ed SIV DNA and RNA occu ed in
he HFD- ecei ing nonhuman p ima es, indica ing a po en ial ese oi expansion. The HFD induced p ominen immune cell
in il a ion in he adipose issue, an impo an SIV ese oi , and heigh ened sys emic immune ac i a ion and in lamma ion,
al e ing he in es inal immune en i onmen and igge ing gu damage and mic obial ansloca ion. Fu he mo e, HFD al e ed
lipid me abolism and HDL oxida ion and also induced li e s ea osis and ib osis. These me abolic dis u bances igge ed
incipien a he oscle osis and heigh ened ca dio ascula isk in he SIV-in ec ed HFD- ecei ing nonhuman p ima es. Ou s udy
demons a es ha die a y in ake has a disce nable impac on he na u al his o y o HIV/SIV in ec ions and sugges s ha
die a y changes can be used as adju an app oaches o HIV-in ec ed subjec s, o educe in lamma ion and he isk o non-
AIDS como bidi ies and possibly o he in ec ious diseases.
High- a die exace ba es SIV pa hogenesis and
accele a es disease p og ession
Tianyu He,1,2 Cuiling Xu,1,3 Noah K ampe1, S ephanie M. Dillon,4 Paola Se e,1,3 Elizabe h Falwell,1,3 Geo ge S. Ha e -Rich e ,1,2
Ti any Bu e ield,5 Tammy L. Dunsmo e,1 William M. McFadden J .,1 Ka h yn J. Ma in,1 Benjamin B. Policicchio,1,6
Ke in D. Raeh z,1,3 Ellen P. Penn,1 Russell P. T acy,7 Ruy M. Ribei o,8,9 Daniel N. F ank,4 Ca a C. Wilson,4 Alan L. Landay,5
C is ian Ape ei,1,3,6 and I ona Pand ea1,2,6
1Cen e o Vaccine Resea ch, 2Depa men o Pa hology, and 3Di ision o In ec ious Diseases, Depa men o Medicine, School o Medicine, Uni e si y o Pi sbu gh, Pi sbu gh, Pennsyl ania, USA. 4Di ision
o In ec ious Diseases, Uni e si y o Colo ado, Anschu z Medical Campus, Au o a, Colo ado, USA. 5Depa men o Mic obial Pa hogens and Immuni y, Rush Uni e si y, Chicago, Illinois, USA. 6Depa men o
In ec ious Diseases and Mic obiology, G adua e School o Public Heal h, Uni e si y o Pi sbu gh, Pi sbu gh, Pennsyl ania, USA. 7Depa men o Pa hology and Labo a o y Medicine, Uni e si y o Ve mon ,
Bu ling on, Ve mon , USA. 8Los Alamos Na ional Labo a o y, Los Alamos, New Mexico, USA. 9Faculdade de Medicina da Uni e sidade de Lisboa, Lisbon, Po ugal.
Con lic o in e es : The au ho s ha e decla ed ha no con lic o in e es exis s.
Copy igh : © 2019, Ame ican Socie y o Clinical In es iga ion.
Submi ed: Ma ch 20, 2018; Accep ed: Sep embe 10, 2019; Published: No embe 11, 2019.
Re e ence in o ma ion: J Clin In es . 2019;129(12):5474–5488.
h ps://doi.o g/10.1172/JCI121208.
The Jou nal o Clinical In es iga ion RESEARCH ARTICLE
5475
jci.o g Volume 129 Numbe 12 Decembe 2019
ec ed NHP species. HFD also signi ican ly inc eased sys emic
immune ac i a ion and in lamma ion and al e ed he gu mucosal
en i onmen , p o oundly comp omising gu in eg i y and inc eas-
ing mic obial ansloca ion. Fu he mo e, HFD induced li e s e-
a osis, ib osis, and al e a ions o he lipid me abolism. Finally,
HFD inc eased equency and se e i y o CV lesions. Al oge he ,
ou esul s p o ide s ong e idence o he c i ical ole o HFD in
modula ing HIV/SIV mo bidi y and mo ali y.
Resul s
S udy design. We used a die ha p o ided 32% calo ies om a
(Table 1), as opposed o no mal chow, which only p o ides 13%
calo ies om a (Table 1). The composi ion o he HFD we used
closely mimicked ha o a ypical Ame ican die o he high con-
en o sa u a ed a s and choles e ol (23).
HFD was adminis e ed o 2 NHP models o nonpa hogenic
(AGM) and pa hogenic (PTM) SIV in ec ion. The goal was no o
compa e he HFD impac be ween species, bu a he o assess he
HFD consequences wi hin he same species, as di ec compa isons
be ween species may be lawed by majo in insic immunologic
and me abolic di e ences. In he same line o eason, we did no
pe ec ly ma ch he iming o HFD adminis a ion in he 2 species.
Fou AGMs ecei ed HFD om 102 days be o e in ec ion and
up o 200 days a e in ec ion (dpi) (Figu e 1). One HFD- ea ed
AGM died du ing he ollow-up. The emaining 3 AGMs we e
eleased a he s udy comple ion. Nine his o ical con ol AGMs
we e used (Figu e 1). They we e SIV in ec ed in simila condi-
ions, had a simila ollow-up and sampling schedule o he
AGMs ecei ing HFD, bu ecei ed a egula chow die (Table 1).
The AGM his o ical con ols we e no subjec ed o in es inal and
li e esec ions.
Fi e PTMs ecei ed HFD om 39 days be o e SIV in ec ion
un il hey p og essed o AIDS (Figu e 1). Ten PTMs we e used as
con ols (Figu e 1). They we e SIV in ec ed in simila condi ions
bu ecei ed a egula chow die (Table 1). Th ee PTMs we e con-
empo a y con ols and had a ollow-up and sampling schedule
simila o ha o HFD- ecei ing PTMs (including in es inal and
li e esec ions). They ei he p og essed o AIDS o we e eu ha-
nized when he las HFD- ecei ing PTM died wi h AIDS. The
emaining 7 PTMs we e his o ical con ols. They had a ollow-up
and sampling schedule simila o ha o HFD- ecei ing PTMs,
bu we e no subjec ed o in es inal and li e esec ions. They we e
eu hanized when hey p og essed o AIDS. All con ols ecei ed
he same chow die .
HFD has a majo impac on disease p og ession and su i al o
SIV-in ec ed NHPs. Fi s , we ques ioned whe he HFD a ec ed key
pa ame e s o HIV/SIV in ec ion (i.e., CD4+ T cell coun s and pVLs).
Pe iphe al CD4+ T cell loss was signi ican ly mo e p ominen
in he SIVsab-in ec ed AGMs ecei ing HFD compa ed wi h con-
ols. As such, al hough a end owa d eco e y was obse ed
du ing ch onic in ec ion, he CD4+ T cell coun s we e s ill sig-
ni ican ly lowe in HFD- ecei ing AGMs compa ed wi h con ols
(Figu e 2A). In e es ingly, a CD4+ T cell dec ease end occu ed
in AGMs a e HFD adminis a ion e en be o e SIV in ec ion,
which was no due o CD4+ T cell down egula ion (24), as no
inc eases in he CD8-αlow o CD4negCD8neg T cells we e obse ed
in he HFD- ea ed AGMs du ing he ollow-up (da a no shown).
Addi ionally, HFD is a classical isk ac o o CV diseases.
HFD induces a he oscle osis (18) by enhancing dyslipidemia and
mac ophage in il a ion in he ao a (18). HFD also al e s he coag-
ula ion p o iles and p omo es hype coagulopa hy, by inc easing
plasma le els o endogenous h ombin and p ocoagulan ac o s
(e.g., FII, FVII, FVIII, and FXII), o dec eased ib inolysis (19).
Hype coagulabili y is common in HIV/SIV in ec ions and may
con ibu e o esidual immune ac i a ion and in lamma ion (20)
e en in indi iduals wi h ART supp ession (2, 3). By al e ing he lip-
id me abolism and coagula ion, HFD may hus enhance he inci-
dence o CV como bidi ies, he mos equen complica ions, and
he leading cause o mo ali y in HIV-in ec ed subjec s (21).
As such, HFD may signi ican ly a ec he na u al his o y o
HIV/SIV, limi ART e icacy, p omo e HIV/SIV- ela ed como bid-
i ies, and ul ima ely inc ease mo ali y (2, 3). Only 1 s udy p e i-
ously epo ed accele a ed disease p og ession in SIV-in ec ed
hesus macaques ecei ing HFD (22). The unde lying mecha-
nisms and he po en ial impac on SIV- ela ed como bidi ies we e
no in es iga ed in de ail, howe e , and only he model o pa ho-
genic HIV in ec ion was e alua ed.
To ho oughly assess he impac o HFD on HIV/SIV na u al
his o y and associa ed como bidi ies, we conduc ed a comp e-
hensi e s udy in which we adminis e ed a HFD ich in sa u a ed
a s and choles e ol o 2 NHP models o SIVsab in ec ion: A i-
can g een monkeys (AGMs), in which in ec ion is nonpa hogenic
(i.e., lacking gu damage, immune ac i a ion and in lamma ion,
como bidi ies and disease p og ession) and pig ailed macaques
(PTMs), in which in ec ion is pa hogenic (i.e., associa ing massi e
gu damage, immune ac i a ion and in lamma ion, como bidi ies,
p og ession o AIDS and dea h).
We epo ha HFD adminis a ion had a majo impac on
i al ese oi seeding and disease p og ession in bo h SIV-in-
Table 1. Compa ison o ND and HFD composi ion
ND HFD
Composi ionA
Fa , % 5.0 15.2
Choles e ol, ppm 80 946
Linoleic acid, % 1.46 3.02
Linolenic acid, % 0.09 0.09
A achidonic acid, % <0.01 0.01
Omega-3 a y acids, % 0.15 0.02
To al sa u a ed a y acids, % 1.36 6.05
To al monounsa u a ed a y acids, % 1.72 4.85
P o ein, % 15.6 18
Ca bohyd a es, % 60 55.6
Calo ies p o ided by
P o ein, % 17.96 16.7
Fa , % 12.95 31.8
Ca bohyd a es, % 69.09 51.5
HFD, high- a die ; ND, no mal die ; ppm, pa s pe million. ANu ien s
a e exp essed as pe cen age o a ion excep whe e o he wise indica ed.
Mois u e con en is assumed o be 10.0% o he pu pose o he calcula ions.
Bolded alues ep esen majo di e ences be ween HFD and no mal chow.
The Jou nal o Clinical In es iga ion
RESEARCH ARTICLE
5476 jci.o g Volume 129 Numbe 12 Decembe 2019
(Figu e 2I). AIDS-de ining his ological lesions also occu ed ea lie
in HFD- ecei ing PTMs (Supplemen al Table 1). A nec opsy, he
HFD- ecei ing PTMs p esen ed wi h mo e lymphoid deple ion,
oppo unis ic in ec ions, and mac ophage in il a ion in issues
compa ed wi h con ols sac i iced a he same ime poin s a e
in ec ion. The CV lesions we e also mo e se e e in he HFD- ecei -
ing PTMs (Supplemen al Table 2). Al oge he , hese ea u es indi-
ca e an ea ly p og ession o AIDS and de elopmen o mo e se e e
SIV- ela ed como bidi ies in he HFD- ea ed animals.
Mos s ikingly, despi e he nonpa hogenic na u e o SIVsab
in ec ion in AGMs, 1 HFD- ecei ing AGM died du ing ollow-up
(Figu e 2J), wi h mul iple bioma ke s p edic ing HIV/SIV disease
p og ession inc easing in he ange ound in PTMs wi h end-s age
disease: mucosal (Figu e 3A) o sys emic (Figu e 3B) T cell ac i a-
ion, LPS (Figu e 3C), hyalu onic acid (Figu e 3D), soluble in e -
cellula adhesion molecule-1 (Figu e 3E); and D-dime (Figu e
3F). A nec opsy, his AGM p esen ed wi h lesions ep esen a i e
o AIDS (oppo unis ic in ec ions) and SIV- ela ed como bidi ies:
en e i is cha ac e ized by a se e e dis up ion o he small in es ine
a chi ec u e wi h epi helial lesions (Figu e 3G) and lamina p op ia
in il a ion wi h oamy cells (Figu e 3H). Such shee s o oamy
mac ophages in he lamina p op ia a e cha ac e is ic o a ypical
mycobac e ia in ec ion o he se e ely immunosupp essed hos s.
Al hough he acid- as s aining could no iden i y he bacilli in he
gu , occu ence o li e g anulomas in his animal also suppo he
diagnosis o a ypical mycobac e ia in ec ion (Figu e 3I).
Se e e in il a i e cys oisospo iasis was also diagnosed in his
AGM (Figu e 3J) and con i med h ough Giemsa s aining (Figu e
3K). The lesions ound in he in es inal epi helium and lamina p o-
p ia a e pa hognomonic o in ec ion wi h Cys oisospo a belli ( o -
HFD- ecei ing AGMs had highe plasma VLs, which eached
signi icance du ing ch onic in ec ion (Figu e 2B), along wi h signi i-
can ly highe PBMC-associa ed SIV DNA le els du ing acu e in ec-
ion (Figu e 2C), sugges ing ha HFD may po en ia e ea ly ese -
oi seeding. This e ec was con i med on he gu samples om
he HFD- ecei ing AGMs, in which acu e p o i al DNA ( DNA)
le els (Figu e 2D), along wi h acu e and ch onic i al RNA ( RNA)
le els (Supplemen al Figu e 1A; supplemen al ma e ial a ailable
online wi h his a icle; h ps://doi.o g/10.1172/JCI121208DS1)
we e signi ican ly highe han in con ols. The acu e RNA le els
we e also highe in he lymph nodes (LNs) o he HFD- ecei ing
AGMs han in con ols (Supplemen al Figu e 1B).
In SIVsab-in ec ed PTMs, which expe ience se e e CD4+ T
cell deple ion h oughou in ec ion, HFD induced u he sig-
ni ican CD4+ T cell loss a la e ime poin s (Figu e 2E). Nei he
pVLs no PBMC o in es ine cell-associa ed DNA le els we e
signi ican ly inc eased in he HFD- ecei ing PTMs compa ed
wi h con ols (Figu e 2, F–H). We de ec ed signi ican ly highe
cell-associa ed RNA le els du ing ch onic in ec ion in bo h in es-
ine (Supplemen al Figu e 1C) and LNs (Supplemen al Figu e 1D)
in he HFD- ecei ing PTMs, howe e , sugges ing highe i al
eplica ion in issues. No e ha SIVsab-in ec ed PTMs unde go
massi e i al eplica ion and yield e y high VLs (25), which likely
masked he HFD e ec on VL.
These da a sugges ha HFD inc eases i us eplica ion, pa icu-
la ly in issues and may hus ha e an impac on he i al ese oi size.
HFD a ec ed SIV su i al in bo h species (Figu e 2, I and J).
Al hough he limi ed sample size p ecluded eaching s a is ical
signi icance, he e was a clea end owa d as e p og ession o
AIDS in HFD- ecei ing PTMs compa ed wi h con ols (P = 0.0522)
Figu e 1. The s udy design o SIVsab: HFD adminis a ion and sampling schedule. A o al o 13 AGMs and 15 PTMs we e in ec ed wi h 300 TCID50 o
SIVsab. Fou AGMs ini ia ed HFD a 102 days be o e in ec ion and main ained i un il 200 days a e in ec ion (dpi). Nine his o ical AGM con ols we e
included and ecei ed no mal chow. Fi e PTMs ini ia ed HFD a 39 days be o e in ec ion and main ained i un il nec opsy (128–218 days a e in ec ion).
Th ee con empo a y and 7 his o ical con ols we e also included and ecei ed no mal chow. Sampling o he AGMs and PTMs in he s udy and con ol
g oups was pe o med as shown.
The Jou nal o Clinical In es iga ion RESEARCH ARTICLE
5477
jci.o g Volume 129 Numbe 12 Decembe 2019
i ied in he colon by immunohis ochemis y (Figu e 3N, uppe pan-
els). The CMV in ec ion in he colon was also con i med h ough
PCR (Supplemen al Figu e 2). La ge, some imes mul inuclea ed
cells wi h i al cy oplasmic o nuclea inclusions su ounded by a
clea halo (Figu e 3N, lowe panels) also indica i e o CMV in ec ion
we e p esen in he lung along wi h in e s i ial in lamma ion (Figu e
3O). CMV in ec ion o he lung was con i med by PCR (Supplemen-
al Figu e 2). The same echnique iden i ied CMV in he spleen and
submandibula LNs (Supplemen al Figu e 2), and he i us could be
quan i ied in blood (a 67 DNA copies/mL o plasma), hus con-
i ming he dissemina ed CMV in ec ion.
Finally, ch onic glome uloneph i is and h ombo ic mic oan-
giopa hy (TMA), ha is, nume ous mic o h ombi in he kidney and
o he issues (Figu e 3P), was also ound in he AGM p og esso .
me ly known as Isospo a belli) and we e iden i ied, as usually done
in a ou ine his opa hological exam, using HE s aining (Figu e
3J). Di e se e olu iona y o ms o he pa asi e we e easily ecog-
nizable in acuoles loca ed in he cy oplasm o he gu epi helial
cells (Figu e 3, J and K). The acuoles con aining he pa asi es
we e mos equen ly loca ed unde he hos cells’ nuclei ha we e
some imes pushed ho izon ally c ea ing he classical T-shape
o m. The pa asi es we e associa ed wi h eosinophil in il a ion in
he lamina p op ia (Figu e 3L). Cys oisospo iasis is an oppo unis-
ic in ec ion ound in pa ien s in ec ed wi h HIV who ha e se e e
immunosupp ession (26).
Fu he mo e, he AGM p og esso also p esen ed wi h coli is
ha associa ed c yp a ophy and in il a ion o all in es inal laye s
wi h mononuclea cells (Figu e 3M). CMV-posi i e cells we e iden-
Figu e 2. HFD accele a es disease p og ession and a ec s su i al o SIV-in ec ed NHPs. Pe iphe al CD4+ T cell deple ion in SIV-in ec ed AGMs (A) and
PTMs (E) a e shown as index o baseline le els and compa ed a key ime poin s o SIV in ec ion wi hin HFD g oup wi h he F iedman es co ec ed o
mul iple compa isons and be ween HFD and con ol g oups wi h K uskal-Wallis es . Plasma i al loads (B), PBMC-associa ed i al DNA (C), and in es-
inal cell-associa ed DNA (D) le els in SIV-in ec ed AGMs, and in SIV-in ec ed PTMs (F–H) we e compa ed a acu e and ch onic in ec ion be ween HFD
and con ol g oups wi h K uskal-Wallis es . Da a a e p esen ed as indi idual alues wi h medians. Sample size (n) and P alues a e p esen ed on g aphs.
BL, baseline (p ein ec ion p e-HFD); Ac, acu e in ec ion; Ch , ch onic in ec ion; Fa , p ein ec ion pos -HFD; Nx, nec opsy. Kaplan–Meie su i al cu es o
SIV-in ec ed PTMs (I) and AGMs (J) and PTMs a e illus a ed. Compa ison be ween su i al cu es o HFD and con ol g oups is pe o med wi h Man el-Cox
me hod. P alues a e p esen ed on g aphs.
The Jou nal o Clinical In es iga ion
RESEARCH ARTICLE
5478 jci.o g Volume 129 Numbe 12 Decembe 2019
Figu e 3. Ele a ed bioma ke s and pa hologic changes o he HFD- ecei ing AGMs ha p og essed o AIDS-like disease. HFD- ecei ing AGM HFD1 de eloped
an AIDS-like disease, wi h dis inc ly ele a ed le els o HLA-DR and CD38 coexp essing CD8+ T cells in he in es ine (A) and pe iphe y (B), LPS (C), hyalu onic acid
(D), sICAM-1 (E), and D-dime (F), as compa ed wi h he emaining 3 AGMs on HFD. Ac, acu e in ec ion; BL, baseline (p ein ec ion p e-HFD); Ch , ch onic in ec ion;
Fa , p ein ec ion pos -HFD. Rep esen a i e H&E images o lesions o AIDS and SIV-associa ed como bidi ies om HFD1. (G) En e i is, wi h dis up ed in es i-
nal a chi ec u e, i egula c yp s, enla ged illi, and al e a ion o he su ace epi helium. (H) Nume ous oamy mac ophages in he lamina p op ia o he small
in es ine sugges i e o ac i e a ypical mycobac e ial disease. (I) Hepa ic g anulomas, sugges i e o a ypical mycobac e ial disease. (J) Cys oisospo a belli p esen
in acuoles in he in es inal epi helium. (K) Giemsa-s aining con i ma ion o small in es ine C. belli in ec ion. (L) Eosinophil in il a ion in he small in es ine,
cha ac e is ic o pa asi ic in ec ions. (M) Coli is, wi h dis up ed in es inal a chi ec u e, a ophic c yp s and mononuclea cell in il a ion o he mucosa, submuco-
sa, and o he muscle laye s. (N) Uppe panels: CMV immunohis ochemis y, wi h posi i e cells in he colon. Lowe panels: Nume ous la ge cells in he lung, wi h
nuclea and cy oplasmic i al inclusions cha ac e is ic o CMV in ec ion (de ail o O). (O) In e s i ial pneumonia, wi h hickened al eola walls due o in il a ion
wi h mononuclea cells. (P) Nume ous mic o h ombi in he kidney glome uli indica i e o TMA (solid a ow). Dis up ed kidney co ex a chi ec u e wi h i egula ,
small ib o ic glome uli, pa hognomonic o ch onic glome uloneph i is (dashed a ow). (Q) Enla ged capilla ies in he kidney pa enchyma indica i e o ch onic
s asis associa ed wi h hea ailu e. (R) Nume ous hemoside in-laden mac ophages in he lung, indica i e o ch onic s asis associa ed wi h hea ailu e. O iginal
magni ica ions: ×200 (G–I, M, and O–R); ×400 (L); ×600 (J, K, and N).

The Jou nal o Clinical In es iga ion RESEARCH ARTICLE
5479
jci.o g Volume 129 Numbe 12 Decembe 2019
TMA occu s in p og essi e HIV/SIV in ec ions, whe e i is associa -
ed wi h inc eased le els o D-dime (3). The AGM p og esso had
D-dime le els in he ange o PTMs wi h end-s age disease (Fig-
u e 3F). Ch onic s asis (cha ac e is ic o conges i e hea ailu e)
was ound in he li e , kidneys (Figu e 3Q), and lungs (Figu e 3R)
and was cha ac e ized by dila ed capilla ies, inc eased blood con-
en , and nume ous hemoside in-laden mac ophages (Figu e 3R).
HFD exace ba es sys emic immune ac i a ion and in lamma ion.
Immune ac i a ion and in lamma ion a e hallma ks o ch onic
pa hogenic HIV/SIV in ec ions (1). As HFD also inc eases in lam-
ma ion independen ly (10), we nex asked whe he HFD and SIV
in ec ion can syne gis ically d i e immune ac i a ion and in lam-
ma ion. While SIVsab-in ec ed AGMs ypically do no exhibi
sys emic immune ac i a ion and in lamma ion (1), he equency
o CD8+ (Figu e 4A) and CD4+ (Supplemen al Figu e 3A) T cells
exp essing HLA-DR+ and CD38+ inc eased immedia ely in AGMs
a e s a ing he HFD, emaining signi ican ly ele a ed h ough-
ou he ollow-up pe iod, wi h he highes le els o CD38+ HLA-
DR+ T cells being obse ed in he AGM p og esso (Figu e 3, A
and B). The Ki-67+ CD4+ T cells we e also signi ican ly inc eased
du ing bo h acu e and ch onic in ec ion (Supplemen al Figu e 3B)
in HFD- ecei ing AGMs compa ed wi h con ols.
SIV-in ec ed PTMs ecei ing HFD simila ly expe ienced a sig-
ni ican inc ease o he HLA-DR+ CD38+ CD8+ T cells du ing bo h
acu e and ch onic in ec ion (Figu e 4D).
The le els o he sys emic in lamma ion bioma ke C eac-
i e p o ein (CRP) we e signi ican ly highe in HFD- ecei ing
SIV-in ec ed AGMs and PTMs compa ed wi h con ols (Figu e 4,
B and E). The le els o RANTES, a bioma ke o mac ophages
and T cell ec ui men a a he oscle o ic lesion si es (27), we e
also signi ican ly ele a ed compa ed wi h con ols in bo h spe-
cies (Figu e 4, C and F).
HFD induces in lamma ion and immune in il a ion in he adi-
pose issue. Adipose issue is an impo an HIV ese oi (28).
P omp ed by ou esul s showing bo h highe le els o cell-
associa ed RNA as well as DNA and inc eased acu e and ch on-
ic immune ac i a ion and in lamma ion in he HFD- ecei ing
NHPs, we u he in es iga ed he in lamma o y s a us o he adi-
pose issue. In SIV-in ec ed PTMs, HFD was associa ed wi h mo e
se e e in lamma o y in il a es in he adipose issue su ounding
c i ical o gans (GI ac and he hea ), compa ed wi h SIV-in ec -
ed con ols. High-deg ee in lamma o y in il a es, ep esen ed by
mononuclea cells, we e obse ed in he pe i oneal a (Figu e 5,
le panel) and he epica dial adipose issue (Figu e 5, igh pan-
el). Se e e in lamma o y in il a es in he epica dium pene a -
ed in he unica ad en i ia o he adjacen co ona y a e ies. In 1
HFD- ea ed PTM, in lamma o y in il a es pene a ed he en i e
co ona y wall, inducing signi ican damage and p omo ing leuko-
cy e adhesion (Figu e 5, igh panel, a ow).
Inc eased me abolic ac i i y may be a sou ce o he immune ac i-
a ion obse ed in HFD- ecei ing NHPs. Inc eased cell me abolism
may d i e immune ac i a ion. Using he seaho se echnology, we
assessed he impac o HFD on he me abolic s a us in PTMs and
showed ha HFD al e ed he PBMC me abolic s a e e en be o e
SIV in ec ion. We quan i ied ae obic glycolysis by moni o ing he
ex acellula acidi ica ion a e (ECAR) a basal espi a ion and in
esponse o oligomycin ( o measu e ATP p oduc ion), luo oca -
bonyl cyanide phenylhyd azone ( o measu e maximal espi a ion),
Figu e 4. HFD induces ele a ed sys emic immune ac i a ion and in lamma ion in SIV-in ec ed NHPs. F equencies o CD38 and HLA-DR coexp essing
CD8+ T cells (A), CRP le els (B) and RANTES le els (C) o baseline le els in AGMs, as well as in PTMs (D–F), a e compa ed a key ime poin s o SIV in ec ion
wi hin HFD g oup wi h F iedman es co ec ed o mul iple compa isons, and be ween HFD and con ol g oups wi h he K uskal-Wallis es . Da a a e p e-
sen ed as indi idual alues wi h medians. Sample size (n) and P alues a e p esen ed on g aphs. Ac, acu e in ec ion; BL, baseline (p ein ec ion p e-HFD);
Ch , ch onic in ec ion; Fa , p ein ec ion pos -HFD; Nx, nec opsy.
The Jou nal o Clinical In es iga ion
RESEARCH ARTICLE
5480 jci.o g Volume 129 Numbe 12 Decembe 2019
and o enone addi ion ( o measu e nonmi ochond ial espi a ion)
(Supplemen al Figu e 4A). PBMCs collec ed a e adminis e ing
he HFD had signi ican ly highe AUC o ECAR compa ed wi h
PBMCs collec ed be o e HFD (Supplemen al Figu e 4B). The
oxygen consump ion a e (OCR) was also measu ed o quan i y
mi ochond ial oxida i e phospho yla ion. A e HFD, PBMCs had
a end o a lowe OCR/ECAR a io a basal espi a ion (Supple-
men al Figu e 4C), indica ing a me abolic shi owa d ae obic
glycolysis. A e HFD, PBMCs also showed mode a ely inc eased
maximal espi a ion OCR, spa e espi a o y capaci y and ATP
p oduc ion, albei wi hou eaching signi icance (P > 0.1) (Supple-
men al Figu e 4D). These esul s collec i ely indica e a pos -HFD
PBMC me abolic shi owa d ae obic glycolysis, e lec ing and
con ibu ing o a heigh ened ac i a ion s a us.
HFD comp omises gu immune in eg i y, p omo ing mic obial
ansloca ion. A majo con ibu o o HIV-/SIV- ela ed immune
ac i a ion and in lamma ion is gu dys unc ion and he conse-
quen mic obial ansloca ion (6). Because he HFD-induced
changes could po en ially dis up he in es inal immune homeo-
s asis and epi helial ba ie in eg i y, we nex assessed HFD
impac on he in es inal immune en i onmen and mic obial
ansloca ion. In HFD- ecei ing AGMs, in es inal CD4+ T cells
we e los e en be o e in ec ion, ye wi hou eaching s a is ical
signi icance (P = 0.4029), p obably due o limi ed sample size
(Figu e 6A). Simila o blood, his dec ease was no due o CD4+ T
cell down egula ion (24), as no inc ease o he mucosal CD8-αlow/
CD4negCD8neg T cells occu ed in he HFD- ea ed AGMs (da a no
shown). CD4+ T cell loss in ol ed all memo y subse s (Figu e 6B).
In es inal CD4+ T cell eco e y was minimal in ch onically SIV-
in ec ed HFD- ecei ing AGMs ( o 22.5% ± 4.5% o baseline le -
els), while being mode a e in con ols ( o 47.5% ± 5.9% o baseline
le els) (Figu e 6A), as epo ed (8). In PTMs, he magni ude o he
SIVsab-induced mucosal CD4+ T cell deple ion obscu ed any di -
e ence be ween HFD- ea ed PTMs and con ols.
Among he mucosal CD4+ T cell subse s, T egs showed a
dec easing end in HFD- ed AGMs e en be o e in ec ion (P =
0.125) (Supplemen al Figu e 5A). In PTMs, mucosal CD4+ T egs
we e signi ican ly deple ed h oughou SIV in ec ion (Supplemen-
al Figu e 5B), being signi ican ly lowe in he HFD- ea ed PTMs
compa ed wi h con ols du ing bo h acu e and ch onic in ec-
ion. T eg dec eases could diminish mucosal ole ance o in es-
inal pa hogens and dis up local in lamma ion in he gu in he
HFD- ecei ing NHPs (29).
HFD also induced inc eased ac i a ion o mucosal immune
cells in HFD- ecei ing PTMs, wi h inc eases o Glu -1+ CD38+
HLA-DR+ CD4+ T cells (Figu e 6C) and mac ophage ac i a ion,
indica ed by he inc eased exp ession o CD80 (Figu e 6D) and
CD86 (Figu e 6E), especially du ing ch onic SIV in ec ion.
Fu he mo e, HFD also signi ican ly inc eased myelope -
oxidase+ neu ophil in il a es associa ed wi h mo e p onounced
epi helial lesions and a highe equency o c yp abscesses in
ch onically SIV-in ec ed PTMs (Figu e 7A), sugges ing g ea e
in es inal ba ie damage and s onge in lamma o y esponses
due o heigh ened mic obial ansloca ion.
Sys emic bioma ke s o mic obial ansloca ion and he asso-
cia ed mucosal damage inc eased pos -HFD adminis a ion.
Plasma LPS le els inc eased in HFD- ecei ing AGMs du ing bo h
acu e and ch onic SIV in ec ion (Figu e 7B) and we e highes in
he AGM p og esso (Figu e 3C). No di e ences in LPS le els we e
obse ed be ween he HFD- ecei ing and con ol PTMs, likely
due o he massi e SIV-induced in es inal damage, o o in insi-
cally ele a ed LPS le els in PTMs (30).
Addi ionally, in es inal a y acid-binding p o ein (I-FABP),
which is eleased by nec o ic en e ocy es, inc eased e en p io SIV
in ec ion and emained inc eased h oughou he ollow-up in he
HFD- ecei ing AGMs (Figu e 7C). I-FABP was also signi ican ly
ele a ed in he HFD- ecei ing PTMs, especially du ing acu e SIV
in ec ion and a he ime o p og ession o AIDS (Figu e 7D). These
esul s demons a e ha HFD exace ba ed in es inal dys unc ion
in bo h NHP models.
HFD induces pa hological al e a ions o he li e . P e ious s ud-
ies showed ha HFD a ec s li e in eg i y. In pa icula , HFD-in-
duced low sys emic endo oxemia is esponsible o NAFLD (31).
In ou s udy, s ea ohepa i is occu ed in AGMs 82 days a e HFD
adminis a ion, p io o SIV in ec ion (Figu e 8A). Li e s ea osis
also occu ed in all he HFD- ecei ing PTMs (Figu e 8A) only 49
days a e HFD (a 9 dpi). Because hese PTMs we e a a e y ea -
ly s age o SIV in ec ion, NAFLD was due o die a he han SIV
in ec ion. Con e sely, NAFLD occu ed in only a small ac ion
o con ols and only du ing a la e s age o SIV in ec ion. The a
d ople in il a es we e di usely dis ibu ed h oughou he hepa -
ic lobules o he HFD- ecei ing PTMs (Figu e 8A), whe eas in con-
ols, hey mos equen ly occu ed a he pe iphe y o he hepa ic
lobule, pa hognomonic o hypoxia. The di e en dis ibu ion o
a accumula ion in he li e poin s o di e en mechanisms o
li e s ea osis in ch onically SIV-in ec ed HFD- ecei ing PTMs
and con ols. NAFLD associa ed mode a e sinusoidal ib osis
ha became signi ican du ing ch onic in ec ion, as his ological-
ly shown by collagen quan i ica ion (Figu e 8B), and con i med
by measu ing plasma hyalu onic acid (32), which signi ican ly
inc eased in HFD- ecei ing AGMs and PTMs compa ed wi h
con ols (Figu e 8, C and D). No ably, he AGM p og esso had
Figu e 5. HFD esul s in immune cell
in il a ion in adipose issue in SIV-
in ec ed PTMs. Rep esen a i e H&E
images o pe i oneal adipose issue
a ound GI ac (le panel) and epica -
dial adipose issue ( igh panels). No e
p ominen lymphocy e in il a ion om
he epica dial adipose issue pene a ing
h ough he co ona y wall indica ed by
he a ows. O iginal magni ica ions: ×200.
The Jou nal o Clinical In es iga ion RESEARCH ARTICLE
5481
jci.o g Volume 129 Numbe 12 Decembe 2019
In he pa hogenic SIV in ec ion o PTMs, which al eady asso-
cia es hype coagulabili y (3), HFD adminis a ion did no u he
inc ease he oxidized HDL o sTF, bu induced signi ican ly highe
FVII le els du ing SIV in ec ion (Figu e 9F). Fu he mo e, opo-
nin-I was signi ican ly ele a ed in 3 ou o 5 ch onically SIV-in ec ed
PTMs on HFD, while emaining unde ec able in con ols (Figu e
9G), di ec ly con i ming ha HFD induced se e e ca diac inju y.
HFD inc eases he equency and se e i y o CV lesions in SIV-
in ec ed PTMs. To assess he equency o CV issue lesions in
HFD- ea ed NHPs, we conduc ed a his opa hological diagnos-
ic on he nec opsy samples collec ed om he PTMs. Al hough
he CV lesions ound in he HFD- ea ed PTMs we e simila o
hose in ch onic SIV-in ec ed con ols, hey we e mo e se e e
and occu ed ea lie (Supplemen al Table 2). CV lesions included
mononuclea cell in il a ion in myoca dium (Figu e 10A) and epi-
ca dium (Figu e 10B), which o med lymphoid agg ega es (Figu e
10C), se e e myoca dial cy olysis (Figu e 10D), myoca dial a y
in il a ion (Figu e 10E), and myoca dial mic o h ombi (Figu e
10F). Two HFD- ecei ing PTMs had incipien ao ic dissec ion
(blood pene a ion unde nea h he endo helium) (Figu e 10G).
These lesions usually e ol e in o ao a aneu ysms ha may up-
u e and induce a al hemo hages. Di use o localized myoca di-
al ib osis also occu ed (Figu e 10H), as well as se e e pe ica dial
ib osis, indica i e o ch onic pe ica di is complica ed wi h colla-
he highes le els o plasma hyalu onic acid (~5.5- old inc ease
om baseline) (Figu e 3D), poin ing o a signi ican ole o HFD-
induced li e pa hology on he ch onic SIV in ec ion ou come.
HFD inc eases he isk o CV como bidi ies in SIV-in ec ed NHPs.
HFD, especially wi h high sa u a ed a , is a majo isk ac o o CV
disease (33). We in es iga ed he e ec s o HFD on he coagula ion
s a us and CV heal h o SIV-in ec ed NHPs, o assess he ela i e con-
ibu ion o die a y habi s o he HIV-associa ed CV como bidi ies.
The choles e ol (Figu e 9A) and oxidized HDL (Figu e 9B)
le els inc eased in HFD- ecei ing, SIV-in ec ed AGMs, while
emaining i ually unchanged in con ols. Oxidized HDL di ec -
ly co ela es wi h accele a ed a he oscle osis (34); he e o e, i s
pos -HFD inc ease in SIV-in ec ed AGMs (which do no de elop
CV disease) indica es ha m ul e ec s o HFD on CV heal h. Solu-
ble issue ac o (sTF) also inc eased in HFD- ecei ing, ch onically
SIV-in ec ed AGMs (Figu e 9C), poin ing o ac i a ion o he ex in-
sic coagula ion pa hway and a highe h ombo ic isk. The inc eased
CV isk was also suppo ed by signi ican ly ele a ed le els o soluble
p-selec in (Figu e 9D) and soluble in e cellula adhesion molecule-1
(sICAM-1) (Figu e 9E) h oughou SIV in ec ion in HFD- ecei ing
AGMs compa ed wi h con ols, demons a ing inc eased pla e-
le and endo helial ac i a ion, along wi h inc eased immune cell
ec ui men o he blood essel walls. The AGM p og esso exhibi -
ed he highes sICAM-1 (Figu e 3E), and D-dime le els (Figu e 3F).
Figu e 6. HFD al e s gu immune en i onmen and ac i a ion s a es o he immune cells. (A) Mucosal CD4+ T cell deple ion in HFD- ecei ing and con ol
AGMs is shown as an index o baseline le els and compa ed a key ime poin s o SIV in ec ion wi hin he HFD g oup wi h F iedman es co ec ed o mul-
iple compa isons, and be ween HFD and con ol g oups wi h K uskal-Wallis es . (B) Mucosal-naï e, cen al memo y, and e ec o memo y CD4+ T cells a e
shown as an index o o al baseline mucosal CD4+ T cell le els and compa ed be o e and a e HFD in p ein ec ion AGMs wi h F iedman es . F equencies
o Glu -1, HLA-DR, and CD38 coexp essing CD4+ T cells (C), as well as CD80-exp essing (D) and CD86-exp essing (E) mac ophages in he in es ine o PTMs
a e compa ed a key ime poin s o SIV in ec ion wi hin HFD g oup wi h F iedman es co ec ed o mul iple compa isons and be ween HFD and con ol
g oups wi h K uskal-Wallis es . Da a a e p esen ed as indi idual alues wi h medians. Sample size (n) and P alues a e p esen ed on g aphs. Ac, acu e
in ec ion; BL, baseline (p ein ec ion p e-HFD); Ch , ch onic in ec ion; Fa , p ein ec ion pos -HFD.
The Jou nal o Clinical In es iga ion
RESEARCH ARTICLE
5482 jci.o g Volume 129 Numbe 12 Decembe 2019
immune ac i a ion and in lamma ion a e being ex ensi ely in es-
iga ed. Thus, in es inal dys unc ion was shown o play a cen al
ole in d i ing disease p og ession (and o be po en ially d i en)
in addi ion o in ec ion- ela ed causes, by en i onmen al ac o s,
such as li es yle and die a y in ake. Li le is known, howe e , abou
he syne gy be ween HIV/SIV in ec ion and en i onmen al ac o s
in d i ing pa hogenesis.
Die e ec s on HIV in ec ion a e di icul o s udy in humans,
due o con ounding ac o s (ART- ela ed dyslipidemia and oxici-
y, alcohol, obacco and ec ea ional d ug use, seden a y li es yle),
and e hical limi a ions (expe imen al al e a ions o die may ha e
unknown e ec s, posing isks o human subjec s; in asi e sampling
needed o assess die e ec s on issues is impossible in humans).
NHPs may be employed o ci cum en hese e hical issues, while
enabling he assessmen o he impac o di e en die a y compo-
nen s on SIV pa hogenesis h ough adminis a ion o pu i ied die s
wi h a speci ic composi ion, s aged in oduc ion o a ious die a y
elemen s, and es ing in a s ic ly con olled en i onmen .
Mans ield e al. p e iously epo ed ha an a he ogenic die
led o accele a ed disease p og ession in hesus macaques and
highligh ed he ole o IL-18 in enhancing he disease onse (22).
gen deposi ion (Figu e 10I), which may es ic ca diac unc ion.
Myoca dial ich ome quan i ica ion showed signi ican inc eased
ib osis o he ca diac muscle in he HFD- ea ed PTMs compa ed
wi h con ols (Figu e 10, J and K). Incipien a he oscle osis lesions
( a y s eaks) in co ona ies, abdominal ao a, and ca o id a e ies
we e mo e equen in he HFD- ea ed PTMs compa ed wi h con-
ols (Figu e 10L). Fa y s eaks we e composed mainly by oamy
cells (mac ophages ha engul ed choles e ol and iglyce ides)
and mononuclea cells, wi h T cells likely accumula ing unde -
nea h he ascula endo helium (Figu e 10L, le ), esul ing in a
hickened blood essel in ima (Figu e 10L, middle). O ganized
mic o h ombi ha igge addi ional leukocy e and pla ele adhe-
sion also occu ed in he HFD- ea ed animals (Figu e 10L, igh ).
Discussion
Al hough he li e span o HIV-in ec ed subjec s d ama ically
imp o ed wi h ART, a highe mo ali y a e han in unin ec ed indi-
iduals s ill pe sis s (2). The e o e, he s udy o non-AIDS como -
bidi ies in long- e m ART-supp essed subjec s has gained ac ion
(2). As p omo e s o non-AIDS como bidi ies and he accele a ed
aging (5), he unde lying mechanism(s) o HIV-/SIV-associa ed
Figu e 7. HFD induces in es inal epi helial ba ie damage and inc eases mic obial ansloca ion. (A) Rep esen a i e images o jejunum s ained o
myelope oxidase (b own) collec ed be o e in ec ion, a 9 days a e in ec ion, and a nec opsy om con empo a y con ols and HFD- ecei ing PTMs. No e
subs an ially inc eased myelope oxidase unde nea h he damaged epi helium and adjacen o he c yp abscess. O iginal magni ica ions: ×200. Quan-
i ica ion o he pe cen a ea o he jejunum posi i e o myelope oxidase is shown in he igh panel. Fold inc ease o LPS (B) and in es inal a y-acid
binding p o ein (C) o baseline le els in AGMs, as well as old inc ease o in es inal a y-acid binding p o ein (D) o baseline le els in PTMs a e compa ed
a key ime poin s o SIV in ec ion wi hin HFD g oup wi h F iedman es co ec ed o mul iple compa isons, and be ween HFD and con ol g oups wi h
K uskal-Wallis es . Da a a e p esen ed as indi idual alues wi h medians. Sample size (n) and P alues a e p esen ed on g aphs. Ac, acu e in ec ion; BL,
baseline (p ein ec ion p e-HFD); Ch , ch onic in ec ion; Fa , p ein ec ion pos -HFD; Nx, nec opsy.