Predict disease progression from T‐cell phenotypes in northern pig‐tailed macaques (Macaca leonina) during SIVmac239 infection

Summary

Macaca leonina (northern pig‐tailed macaques, NPMs) have variable disease progression during SIVmac239 infection. In the present study, we analysed, for the first time, the correlations between T‐cell phenotypes and disease progression in NPMs during SIVmac239 infection. In comparison to normal progressors (NPs), slow progressors (SPs) had lower chronic T‐cell activation and exhaustion levels. In addition, SPs showed higher peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio, and lower plasma viral load than NPs. CD4⁺ T‐cell count and CD4 : CD8 ratio decreased more sharply in NPs than in SPs. Furthermore, T cells in NPs were more highly differentiated, at least in acute infection, than in SPs. These results indicated that T‐cell phenotypes were correlated with disease progression in SIVmac239‐infected NPMs and these correlations may provide valuable guidance for the improvement of therapeutic strategies tested in NPMs.

Abbreviations

HIV

human immunodeficiency virus

NPMs

northern pig‐tailed macaques

NPs

normal progressors

PD‐1

programmed cell death protein 1

SHIV

simian‐human immunodeficiency virus

SIV

simian immunodeficiency virus

SPs

slow progressors

Tcm

central memory T

Introduction

Elevated and persistent immune activation is one of the major characteristics induced by human immunodeficiency virus type 1 (HIV‐1) infection.1, 2 Although the cause of immune activation during HIV‐1 infection is not fully understood, persistent immune activation is associated with a poor immunological reconstitution and prognosis.3, 4 It has long been proved that high CD38 or HLA‐DR expression on CD4⁺ and CD8⁺ T cells increases morbidity and mortality in HIV disease.5, 6, 7, 8 In addition, we also found a correlation between T‐cell activation and death of simian/human immunodeficiency virus (SHIV) 89.6‐infected Chinese rhesus macaques.9 Programmed cell death protein 1 (PD‐1), a negative regulator of T‐cell activation and biomarker of T‐cell exhaustion is also proved to be associated with disease progression.10 Expression of PD‐1 on CD4⁺ and CD8⁺ T cells correlates positively with plasma viral load and negatively with CD4⁺ T‐cell count.10, 11 In addition, PD‐1 expression on CD4⁺ and CD8⁺ T cells is associated with their activation respectively11.

Progressive depletion of CD4⁺ T cells induced by direct cytotoxicity or other indirect mechanisms is a hallmark of HIV‐1 infection.12 Immune deficiency and the following opportunistic infections are directly associated with low CD4⁺ T‐cell counts.12 Peripheral CD4⁺ T‐cell count together with plasma viral load count were used as predictors of AIDS disease progression in earlier years, although they were proved not as predictive as immune activation later.6, 13, 14, 15 Nowadays, peripheral CD4⁺ T‐cell count and plasma viral load are still selected as significant indicators of clinical disease progression.16, 17 Besides, the CD4 : CD8 ratio is another easily available parameter that is associated with immunological dysfunction and disease progression.18, 19, 20 A persistently low CD4 : CD8 ratio is associated with increased innate and adaptive immune activation and higher risk of morbidity and mortality.20 Except for the parameters mentioned above, pre‐infection level and loss of CD4⁺ central memory (Tcm) cell may also predict rapid disease progression in Macaca nemestrina.21 Therefore, T‐cell phenotypes and counts may reflect the disease progression to some extent.

Pig‐tailed macaques are valuable HIV/AIDS animal models. Therein, Macaca nemestrina (southern pig‐tailed macaques) are widely used in HIV/AIDS research, while Macaca leonina (northern pig‐tailed macaques, NPMs) are scarcely studied. To establish appropriate HIV/AIDS animal models using NPMs, we infected six NPMs with simian immunodeficiency virus (SIV) mac239, during which NPMs had variable disease progression. In consideration of the evaluation of therapeutic strategies in NPMs, it is significant to study the factors correlated with disease progression. As T‐cell phenotypes were routinely analysed during animal experiments, we decided to study the relationships between disease progression and T‐cell phenotypes, including T‐cell activation and exhaustion, CD4⁺ T‐cell count, CD4 : CD8 ratio and T‐cell differentiation.

Materials and methods

Animals and sample collection

Six male adult (6–8 years old) NPMs were enrolled in our study. They were healthy and negative for SIV, simian type‐D retrovirus, simian T‐lymphotropic virus and B virus before this study. A tissue culture infectious dose 50% of 3000 SIVmac239 was inoculated into each macaque by intravenous injection. All the macaques were housed in Kunming Primate Research Centre in accordance with the guide of the American Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). All the animal experiment procedures in our study were approved by the Ethics Committee of Kunming Institute of Zoology, Chinese Academy of Sciences (approval number: SYDW‐2015023; approval date: 20 June 2015).

Ketamine hydrochloride was used for the anaesthesia of animals. After complete anaesthesia, peripheral blood was collected through venepuncture and blood was drawn into vacuum tubes containing EDTA. Anaesthetized animals were taken good care of until they awoke.

Plasma viral load

Plasma was isolated by 1000 g centrifugation from peripheral blood and cryopreserved in small aliquots in −80°C. Viral RNA was extracted from plasma using a High Pure Viral RNA Kit (Roche Diagnostics GmbH, Roche Applied Science, Mannheim Germany)22 according to the manufacturer's instructions. Plasma viral loads were determined by real‐time PCR (ABI ViiA™ 7 Real‐Time PCR System; Applied Biosystems, Foster City, CA, USA). Primers, probe and PCR conditions were all the same as previously described.9

Multiparameter flow cytometry

Multiparameter flow cytometry was performed using peripheral blood rather than cryopreserved peripheral blood mononuclear cells. The following anti‐human flow cytometry antibodies cross‐reactive with macaques were used: anti‐CD3‐phycoerythrin (PE) (clone SP34‐2), anti‐CD8‐PE/Cy7 (clone RPA‐T8), anti‐CD3‐allophycocyanin/Cy7 (clone SP34‐2) and anti‐HLA‐DR‐allophycocyanin (clone G46‐6) were purchased from BD Pharmingen, Franklin Lakes, New Jersey, USA; anti‐CD4‐Peridinin chlorophyll protein/Cy5.5 (clone OKT4) was from BioLegend (San Diego, CA, USA); anti‐PD‐1‐PE (clone eBioJ105) was from eBioscience (San Diego, CA, USA). Flow cytometric procedures were described previously.23, 24 Flow cytometric acquisitions were performed on a BD FACSVerse™ (Franklin Lakes, New Jersey, USA) flow cytometer and all flow cytometric data were analysed on flowjo software (vX.0.7; TreeStar, Ashland, OR, USA).

Statistical analysis

Two‐way analysis of variance compared the difference of T‐cell activation/exhaustion, peripheral CD4⁺ T‐cell count, CD4 : CD8 ratio and plasma viral load between normal and slow progressor NPMs. P < 0·05 was considered statistically significant. All data analyses were performed using the graphpad prism software (version 6.01; GraphPad Software, La Jolla, CA, USA).

Results

The disease progression varies between SIVmac239‐infected NPMs

Two individuals (08247 and 09211) showed a wasting syndrome (weight loss > 10% versus day 0 for at least 1 month) among six SIVmac239‐infected NPMs (Table 1).25 They died on day 491 and day 476 post infection, respectively, and showed obvious pathological changes in tissues. Two NPMs (08247 and 08287) showed thrombocytopenia during the infection. The other three NPMs (09203, 10205 and 10225) did not show wasting syndrome or thrombocytopenia. Based on the AIDS symptoms, we divided the six NPMs into two groups: normal progressors (NPs) and slow progressors (SPs).

Table 1.

Disease progression status of simian immunodeficiency virus mac239‐infected northern pig‐tailed macaques

Group	Macaque	Death time (days post infection)	AIDS symptoms
Normal progressors	08247	491	Wasting syndromea, thrombocytopeniab, lymphadenopathy, lung lesions, liver lesions
	08287	NDc	Thrombocytopenia
	09211	476	Wasting syndrome, lung lesions
Slow progressors	09203	ND	No wasting syndrome, no thrombocytopenia, alive
	10205	ND
	10225	ND

^aWasting syndrome is defined as weight loss > 10% for at least 1 month.

^bThrombocytopenia is defined as < 100 × 10⁹ platelets/l blood.

^cND, not dead.

Chronic T‐cell activation and exhaustion are associated with disease progression

We analysed T‐cell activation (defined by HLA‐DR expression) and exhaustion (defined by PD‐1 expression) in a long‐term infection. The infection was divided into acute phase and chronic phase on the 12th week post infection. Expressions/co‐expressions of these markers showed similar levels in NPs and SPs in acute phase (Fig. 1). These results implied that acute T‐cell activation and exhaustion might not affect disease progression. Conversely, chronic T‐cell activation and exhaustion levels in NPs were higher than in SPs, with statistical significance, except CD8⁺ T‐cell exhaustion (P = 0·6158) (Fig. 1b). Overall, there were strong relationships between chronic T‐cell activation/exhaustion and disease progression.

Figure 1.

Comparisons of T‐cell activation and exhaustion levels between normal progressor (NP) and slow progressor (SP) northern pig‐tailed macaques (NPMs). Longitudinal analysis of CD38, HLA‐DR and PD‐1 expression/co‐expression on (a) CD4⁺ T cells and (b) CD8⁺ T cells. The infection was divided into acute phase and chronic phase at 12 weeks post infection. Data are shown as median and error. Two‐way analysis of variance compared the difference of activation and exhaustion marker expressions between normal and slow progressor NPMs. P < 0·05 was considered as statistically significant.[Colour figure can be viewed at wileyonlinelibrary.com]

Disease progression was correlated with peripheral CD4⁺ T‐cell count, CD4 : CD8 ratio and plasma viral load

Peripheral CD4⁺ T‐cell count and plasma viral load are clinically conventional surrogate markers of HIV‐1 disease progression.26 In addition, it has been reported that low CD4 : CD8 ratio predicts disease progression.19, 20 Therefore, we analysed the relationships between disease progression and peripheral CD4⁺ T cell, CD4 : CD8 ratio and plasma viral load in our study. Peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio were higher in SPs than in NPs throughout the infection whereas plasma viral load was lower in SPs than in NPs during chronic infection (Fig. 2). These results indicated that peripheral CD4⁺ T‐cell count, CD4 : CD8 ratio and plasma viral load might be appropriate surrogate markers of disease progression.

Figure 2.

Comparisons of peripheral CD4⁺ T‐cell count, CD4 : CD8 ratio and plasma viral load between normal progressor (NP) and slow progressor (SP) northern pig‐tailed macaques (NPMs). Longitudinal analysis of (a) peripheral CD4⁺ T‐cell count, (b) CD4 : CD8 ratio and (c) plasma viral load. Data are shown as median and error. The infection was divided into acute phase and chronic phase at 12 weeks post infection. Two‐way analysis of variance compared the difference of peripheral CD4⁺ T‐cell count, CD4 : CD8 ratio and plasma viral load between normal and slow progressor NPMs. P < 0·05 was considered as statistically significant. [Colour figure can be viewed at wileyonlinelibrary.com]

Disease progression was correlated with fold changes of peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio

Given the individual differences of pre‐infection peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio, we further compared the fold changes of these parameters in different groups. Peripheral CD4⁺ T‐cell count decreased more sharply in NPs than in SPs in both acute and chronic phases (Fig. 3a). In addition, peripheral CD4⁺ T‐cell count in SPs recovered to pre‐infection levels, but not in NPs. The CD4 : CD8 ratio also decreased more severely in NPs than in SPs throughout the infection (Fig. 3b). These results further implied that peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio were potential surrogate markers of disease progression.

Figure 3.

Fold changes of peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio between normal progressor (NP) and slow progressor (SP) northern pig‐tailed macaques (NPMs). Longitudinal analysis of fold changes of (a) peripheral CD4⁺ T‐cell count and (b) CD4 : CD8 ratio. Fold change of day $\mathrm{X}=\frac{\text{Day x}-\text{Day 0}}{\text{Day 0}}\times 100\%$. Data are shown as median and error. The infection was divided into acute phase and chronic phase at 12 weeks post infection. Two‐way analysis of variance compared the fold changes of peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio between normal and slow progressor NPMs. P < 0·05 was considered as statistically significant. [Colour figure can be viewed at wileyonlinelibrary.com]

High differentiation of T cells may predict rapid disease progression

It has been proved that loss of CD4⁺ Tcm cells may predict rapid disease progression in Macaca nemestrina.21 Whether it is the same in Macaca leonina is still unclear. Therefore, we compared the frequencies of T‐cell subsets in different groups. There was no obvious difference in frequencies of CD4⁺ naive T cells (Fig. 4a). But, the frequency of CD4⁺ Tcm cells in SPs was higher than in NPs during acute infection, whereas the frequency of CD4⁺ T effector memory cells was just the opposite. The differences no longer lasted during chronic infection. Although there was no significant difference in the frequencies of CD4⁺ Tcm cells, we could still infer the more severe loss of CD4⁺ Tcm cells in NPs from the CD4⁺ T‐cell count and fold change. The frequency of CD8⁺ naive T cells in SPs was higher than in NPs throughout the infection, whereas the frequency of CD8⁺ T effector memory cells was just the opposite (Fig. 4b). The frequency of CD8⁺ Tcm cells in SPs was higher than in NPs during acute infection, but not in chronic infection. These results indicated that high differentiation of T cells, at least in acute infection, might predict rapid disease progression.

Figure 4.

Frequencies of peripheral CD4⁺ T‐cell subsets. CD4⁺ T cells were divided into naive T cells (CD28⁺ CD95⁻), central memory T cells (CD28⁺ CD95⁺) and effector memory T cells (CD28⁻ CD95⁺). Longitudinal analysis of frequencies of (a) CD4⁺ T‐cell subsets and (b) CD8⁺ T‐cell subsets. Data are shown as median and error. The infection was divided into acute phase and chronic phase at 12 weeks post infection. Two‐way analysis of variance compared the frequencies of CD4⁺ T‐cell subsets and CD8⁺ T‐cell subsets between normal progressor (NP) and slow progressor (SP) northern pig‐tailed macaques. P < 0·05 was considered as statistically significant. [Colour figure can be viewed at wileyonlinelibrary.com]

Discussion

Predictors of disease progression may provide valuable guidance for the improvement of therapeutic strategies tested in HIV/AIDS animal models. Here we studied the relationships between disease progression and various T‐cell phenotypes in NPMs during SIVmac239 infection. Overall, NPs showed higher chronic T‐cell activation and exhaustion levels than SPs. Meanwhile, SPs showed higher peripheral CD4⁺ T‐cell count and CD4 : CD8 ratio, and lower plasma viral load. In addition, T cells were more highly differentiated in NPs than in SPs.

Although CD4⁺ T‐cell counts in SPs were higher and viral loads in SPs were lower than that of NPs, we still cannot eliminate the irregularity of the results. Hence, we further analysed the correlations between T‐cell activation/exhaustion and CD4⁺ T‐cell count, CD4 : CD8 ratio and plasma viral load (see Supplementary material, Fig. [Link], [Link], [Link]). Overall, T‐cell activation/exhaustion is negatively correlated with CD4⁺ T‐cell count and CD4 : CD8 ratio and positively correlated with plasma viral load. But, the correlations are not consistent in different T‐cell subsets or different infection phases. As T‐cell activation has been proved to be more predictive,6, 15 CD4⁺ T‐cell count and viral load may not be as prognostic as the results show. In addition, the correlations between CD4 : CD8 ratio and T‐cell exhaustion are more consistent than that of CD4⁺ T‐cell count and viral load (see Supplementary material, Fig. [Link], [Link], [Link]), which indicates that CD4 : CD8 is more prognostic, as previously reported.20

Although chronic T‐cell activation/exhaustion were proved to be correlated with rapid disease progression in SIVmac239‐infected NPMs, T‐cell activation and exhaustion are complicated states during which a multitude of markers are increasingly expressed.27, 28, 29 More markers may be needed to show the accurate correlations between T‐cell activation/exhaustion and disease progression.

As disease progression is controlled by many factors, more relative immunological parameters may be better for the estimation of disease progression. Thrombocytopenia is used as a factor for the classification of different progression groups, but we also compared the fold change of platelets during SIV infection. The results indicated that a sharp decrease of platelets in chronic infection might predict rapid disease progression in SIVmac239‐infected NPMs (see Supplementary material, Fig. S2). In addition, to precisely prognosticate the disease progression, more predictors such as coagulation biomarkers, microbial translocation‐related factors, thrombocytopenia and plasma cytokine levels may provide complements to future studies.30, 31, 32, 33

Unfortunately, there are also some limitations in our study. First, the small number of NPMs enrolled in our study may not fully reflect the relationships between T‐cell phenotypes and disease progression. Second, to thoroughly analyse the correlations between disease progression and T‐cell phenotypes, more phenotypes other than activation/exhaustion and differentiation should be involved and a comprehensive analysis may provide a profound insight into the correlations. Finally, as large numbers of HIV‐infected patients are receiving treatment, the correlations between T‐cell phenotypes and disease progression in treated NPMs may provide more valuable suggestions for the improvement of therapeutic strategyies.

In summary, we investigated the correlations between disease progression and T‐cell phenotypes in SIVmac239‐infected NPMs. The results indicated that T‐cell phenotypes might predict disease progression to some extent. When therapeutic strategies were tested in NPMs, T‐cell phenotypes might be potentially appropriate indicators of effectiveness.

Disclosures

The authors participating in this study declare that there are no conflicts of interest regarding the publication of this article.

Supporting information

Figure S1. Correlations between T‐cell activation/exhaustion and (a) peripheral CD4⁺ T‐cell count, (b) plasma viral load and (c) CD4 : CD8 ratio.

 

 

Figure S2. Fold change of platelets.