Edited by: Aurelio Cafaro, National Institute of Health (ISS), Italy
Reviewed by: Scott Kitchen, University of California, Los Angeles, United States; Larisa Y. Poluektova, University of Nebraska Medical Center, United States; Yong-Guang Yang, Jilin University, China
*Correspondence: Todd M. Allen,
†This author passed away before completion of this work but made substantial contributions
‡These authors have contributed equally to this work
This article was submitted to Viral Immunology, a section of the journal Frontiers in Immunology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Humanized bone marrow-liver-thymus (HuBLT) mice are a revolutionary small-animal model that has facilitated the study of human immune function and human-restricted pathogens, including human immunodeficiency virus type 1 (HIV-1). These mice recapitulate many aspects of acute and chronic HIV-1 infection, but exhibit weak and variable T-cell responses when challenged with HIV-1, hindering our ability to confidently detect HIV-1–specific responses or vaccine effects. To identify the cause of this, we comprehensively analyzed T-cell development, diversity, and function in HuBLT mice. We found that virtually all HuBLT were well-reconstituted with T cells and had intact TCRβ sequence diversity, thymic development, and differentiation to memory and effector cells. However, there was poor CD4+ and CD8+ T-cell responsiveness to physiologic stimuli and decreased TH1 polarization that correlated with deficient reconstitution of innate immune cells, in particular monocytes. HIV-1 infection of HuBLT mice showed that mice with higher monocyte reconstitution exhibited greater CD8+ T cells responses and HIV-1 viral evolution within predicted HLA-restricted epitopes. Thus, T-cell responses to immune challenges are blunted in HuBLT mice due to a deficiency of innate immune cells, and future efforts to improve the model for HIV-1 immune response and vaccine studies need to be aimed at restoring innate immune reconstitution.
Human immunodeficiency virus type 1 (HIV-1) first arose in Africa as a cross-species transmission event of simian immunodeficiency virus (SIV) in the 1930s (
Several humanized mouse models have been developed, mostly stemming from the introduction of severely immunodeficient mouse strains capable of achieving high levels of reconstitution with human cells. One widely used strain is NSG mice, which are non-obese diabetic (NOD) mice bearing
This barrier of MHC restriction was overcome with the development of HuBLT mice (
In this study, we comprehensively analyzed T-cell development, diversity, and function in HuBLT mice to identify barriers that explained deficiencies in immune responses to HIV-1 infection from that of adult humans. We found that while T-cell development and diversity is intact in HuBLT mice, there is a defect in T-cell function and responses to HIV-1 infection that correlates strongly with limited innate immune reconstitution. Thus, this study highlights that innate immune reconstitution is likely a major barrier to normal T-cell responses to HIV-1 infection in HuBLT mice.
NOD-
Isolated leukocytes were stimulated by incubating for 4 – 6 h with phorbol 12-myristate 13-acetate (PMA) (12.5 ng/mL) and ionocymin (0.335 μM) (Cell Stimulation Cocktail used at 0.25X; eBioscience) or anti-CD3/28 Dynabeads (Life Technologies) at a bead-to-cell ratio of approximately 2:1 in the presence of brefeldin A (BioLegend) and monensin (BD Biosciences) at the manufacturers recommended concentrations. Cells were then subsequently stained with fluorescently labeled antibodies and assessed
Direct staining of peripheral blood leukocytes was performed by addition of fluorescently labeled antibodies to whole blood and performing RBC lysis and fixation with BD FACS Lysing Solution (BD Biosciences). For stimulation experiments, peripheral blood leukocytes were isolated from whole blood by density gradient centrifugation with Histopaque (Sigma-Aldrich) to generate a layer of live mononuclear cells that was collected and washed with cell culture media consisting of 10% fetal bovine serum (Sigma-Aldrich), L-glutamine (Corning), and Primocin antibiotic (Invivogen) in RPMI-1640 (Thermo Fisher). Cells from tissues (e.g. spleen, thymus) were mechanically extracted by placing the tissue sample in a 70-μm cell strainer (Corning) in a well of a 6-well plate containing ~5 mL of cell culture media, and mashing carefully but firmly against the strainer mesh until the tissue was dissociated. The single cell suspension then underwent density gradient centrifugation to isolate live mononuclear cells.
Surface staining was performed by incubating with the corresponding antibodies at 4°C for 15 min. Cells were then washed with 2% FBS and 2 mM EDTA in PBS and fixed with 4% paraformaldehyde in PBS (Affymetrix). In experiments where CD107a surface expression was measured, the corresponding antibody was pre-incubated with the cells during stimulation for optimal staining. For measurement of cytokine-producing cells after stimulation, intracellular staining for cytokines was performed by using the BD Cytofix/Cytoperm fixation/permeabilization kit (BD Biosciences) following the manufacturer’s protocol and staining with the corresponding surface and intracellular antibodies as instructed. When applicable, anti-CD3 antibody was included in the intracellular stain to increase CD3 staining given that stimulation results in partial CD3 downregulation. Intranuclear staining for TdT was also performed by using the BD Cytofix/Cytoperm fixation/permeabilization kit. See
Flow cytometry data was acquired on BD LSR Fortessa and analyzed using FlowJo software (version 10), and statistical analyses were performed using Microsoft Office Excel, JMP Pro 14, and GraphPad Prism 8.
RNA was extracted from isolated leukocytes using the RNeasy Plus Mini Kit (Qiagen) and QIAshredder Kit (Qiagen) following manufacturer’s instructions. 5’ rapid amplification of cDNA ends (5’ RACE) was then performed using the SMARTer RACE cDNA Amplification Kit (Clontech). cDNA was then amplified with a first round of 5’ RACE PCR using the Advantage-HF 2 Polymerase Mix (Clontech) with a 5’ universal primer mix (provided by the kit) and a gene-specific primer that recognizes all constant regions of TCRβ:
TCRβout (5’→3’): TGTGGCCAGGCACACCAGTGTGGCC
A follow-up nested PCR was performed using a nested universal primer containing an adaptor for 454 pyrosequencing (NUP) and a nested gene-specific primer that recognized all TCRβ constant regions and contained an adaptor for 454 pyrosequencing and a barcode (TCRβin):
NUP (5’→3’): CCTATCCCCTGTGTGCCTTGGCAGTCTCAGCAAGCAGTGGTATCAACGCAGAG
TCRβin (5’→3’): CCATCTCATCCCTGCGTGTCTCCGACTCAG(N)10GCTCAAACACAGCGACCTCGGGTGGGA
where (N)10 is barcoded region.
Gel band extraction for a band of approximately 450 – 500 bp was then performed using Purelink Quick Gel Extraction Kit (Invitrogen). PCR purification with QIAquick PCR Purification Kit was performed, and DNA was quantified using QUANTI-IT PicoGreen dsDNA Reagent Assay (Invitrogen) and fluorometer (Promega). Pooled PCR products were prepared for sequencing on the 454 Genome Sequencer FLX Titanium (Roche) using standard protocols (specifically, Lib L kit) and following manufacturer’s instructions. Sequence reads were analyzed using the IMGT/HighV-QUEST tool (
Mouse tissues were freshly extracted and placed into 4% paraformaldehyde in PBS (Affymetrix) for 48 h at 4°C, and then stored in 70% ethanol at 4°C until being sent to MGH Histopathology Research Core for embedding in paraffin, sectioning, and immunofluorescent and immunohistochemical staining. Stained tissue was visualized on a TissueFAXS (TissueGnostics).
Amplification and next-generation deep sequencing of viral genomes derived from HuBLT mice has been previously described for this data set in detail (
TCR sequencing analyses are available at VDJServer (UUID: 3958818965646011925-242ac116-0001-012;
In order to characterize thymic development, thymocytes were extracted from thymic organoids and stained with human markers of T-cell development for flow cytometric analysis (
Human thymic organoids in HuBLT mice sustain thymopoiesis.
Given existing literature suggesting that fetal T-cell repertoires are limited in diversity (
TCR diversity in HuBLT mice is comparable to adult humans.
To further investigate the extent of T-cell development and differentiation, we assessed the frequency of CD8+ versus CD4+ and naïve versus memory T cells. Flow cytometric analyses of the peripheral blood of HuBLT mice generated from different tissue donors at different times post-engraftment demonstrated that CD4+ T cells are often the most abundant human cells in peripheral blood of HuBLT mice, with very high CD4:CD8 T-cell ratios (
CD4/CD8 T-cell ratios and naïve T cell frequencies are increased in HuBLT mice but change with age.
Phenotypic analysis of naïve and memory T-cell subsets was performed by staining for CD45RA and CCR7. Peripheral blood T cells of HuBLT mice contained naïve (Tnaïve; CD45RA+CCR7+), central memory (TCM; CD45RA–CCR7+), effector memory (TEM; CD45RA–CCR7–), and effector memory re-expressing CD45RA (TEMRA; CD45RA+CCR7–) T-cell subsets that varied within the limits of healthy adult human peripheral blood (
Having confirmed that the development and differentiation of T cells was intact, we sought to assess their function in response to a variety of stimuli and immune challenges. We stimulated bulk peripheral blood leukocytes with phorbol 12-myristate 13-acetate and ionomycin (P+I), which mimic diacylglycerol and calcium release downstream of TCR and co-receptor signaling, and performed surface and intracellular staining. This demonstrated that P+I potently induced IL-2 production in CD4+ T cells, degranulation of CD8+ T cells, and IFN-γ production from both (
T cells in HuBLT mice have impaired responsiveness that correlates with inadequate innate immune reconstitution.
Interestingly, CD4+ T cells that produced IFN-γ in response to P+I stimulation were found to be lower than what would be expected for adult human peripheral leukocytes (
TH1 polarization in HuBLT mice is impaired but improves with innate immune reconstitution and age.
Studies have reported that chemokine receptor expression can be used as a surrogate for TH polarization status, namely, that CXCR3 expression correlates with IFN-γ production (TH1 polarization) (
To assess whether T-cell responsiveness was affected by innate immune reconstitution
CD8+ T cell responses to HIV-1 infection correlate with innate immune reconstitution.
In order to draw a link between innate immune reconstitution and pathogen-targeted immune responses, we evaluated the effect of monocyte reconstitution on HIV-1 sequence diversity within predicted HLA-restricted epitopes as a metric of CD8+ T cell-mediated immune pressure (
The requirement of innate antigen-presenting cells for maintaining and priming T-cell immunity is a well-known immunological concept that has been re-emphasized by our comprehensive analysis of T-cell immunology in HuBLT mice. We find that the model achieves the desired endogenous thymic development, TCR repertoire diversity, and generation of T-cell subsets comparable to that of humans, suggesting that efforts at improving these processes are likely not required in the model. However, poor reconstitution of monocytes significantly correlated to defects in T-cell function across multiple contexts, including functional maintenance, TH polarization, priming, and anti-viral responses during HIV-1 infection. The latter finding was a striking observation that monocyte reconstitution not only significantly correlated with CD8+ T cell activation, but also with CD8+ T cell-mediated immune pressure as measured by viral evolution within (but not outside) HLA-restricted epitopes, suggesting that human-like HIV-1-specific T-cell responses can arise in HuBLT mice when all necessary elements are present. Prior experiments in immunocompetent mice have demonstrated that
Defective reconstitution of innate immune cells has been previously described in humanized mouse models (
Cross-reactivity of human and mouse cytokines and their receptors.
Cytokine | Cross-reactivity (ligand → receptor) | Relevant Function | |
---|---|---|---|
mouse → human | human → mouse | ||
|
↓2.5-fold | none | Myelopoiesis |
|
none | none | Myelopoiesis |
|
none | none | Myelopoiesis |
|
none | full | Myelopoiesis |
|
active | full | Myelopoiesis |
|
↓2.5-fold | active | Myelopoiesis |
|
full | active | Myelopoiesis |
|
full | full | Myelo-/Lymphopoiesis |
|
none | active | Myelo-/Lymphopoiesis |
|
active | active | Lymphopoiesis |
|
none | active | Lymphopoiesis |
|
none | none | Lymphopoiesis |
|
active | active | Lymphopoiesis |
|
active | none | Lymphopoiesis |
|
active | active (not TNFR2) | Lymphopoiesis |
|
full | full | Lymphopoiesis |
|
active | active | Lymphopoiesis, |
|
full | (n.d.) | T-cell zone homing |
|
↓100-fold | (n.d.) | B-cell zone homing |
|
none | none | Lymph node genesis |
‘full’ = approximately 100% cross-reactivity has been reported; ‘active’ = cross-reactivity has been reported but not quantitated; ‘none’ = negligible activity has been reported; ‘(n.d.)’ = no data readily available.
Non-cross-reactivity of these critical factors has led many investigators to develop strategies that introduce human cytokines and growth factors by various methods including (
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: NCBI BioProject, accession no: PRJNA552879.
The animal study was reviewed and approved by Massachusetts General Hospital Institutional Animal Care and Use Committee.
WG-B, DC, CM, MK, MA, and TA designed experiments. WG-B, DC, CM, MP, VV, MK, KK, and KP conducted experiments. KP, CB, AT, MA, and TA provided significant and insightful contributions to design of experiments. VV, AT, MA, and TA provided significant resources. WG-B and DC wrote the paper with contributions from all other authors. All authors contributed to the article and approved the submitted version.
This work was supported by the US National Institute of Health (P01-AI104715 to MA, and TA; F31AI116366 to WG-B; 1F32AI136750 to DC; 5T32AI007529-21A1 to MK), the National Institute of General Medical Sciences (T32GM007753 to WG-B), and the Ragon Institute of MGH, MIT and Harvard. AB is supported by the National Institutes for Drug Abuse (NIDA) Avenir New Innovator Award DP2DA040254, the MGH Transformative Scholars Program as well as funding from the Charles H. Hood Foundation. This independent research was supported by the Gilead Sciences Research Scholars Program in HIV. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.
The content is solely the responsibility of the authors and does not necessarily represent the official views or policies of the National Institute of General Medical Sciences, the National Institutes of Health, or the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The Supplementary Material for this article can be found online at: