Background
The insulin-like growth factor (IGF) system consists of growth peptides (IGF1 and IGF2), their receptors (IGF1R and IGF2R), and circulating binding proteins (IGFBPs 1-6) [
1-
3]. Multiple factors influence IGF levels in biological fluids including nutrition, endocrine status, liver disease, chronic inflammatory diseases, diabetes mellitus, and ethnicity (African American < Asian < Caucasian) [
4]. Plasma levels of IGFs are much higher (10 to 100 fold) than required for biological action, as the majority are bound to IGF-binding proteins (IGFBPs) [
5,
6]. Most circulating IGF1 is produced by the liver, which is under the regulation of pituitary growth hormone (GH). Extrahepatic (non-GH-dependent) IGF1 synthesis also occurs [
3]. IGF2 is also produced from hepatic and extrahepatic sources but its production is not tightly regulated by GH [
7]. IGF1 and IGF2 are also synthesized in the CNS and surrounding cells [
8-
10] and have neurotrophic effects [
11-
14]. Both IGF1 and IGF2 signal through IGF1R, whereas IGF2R is a non-signaling receptor involved in the removal of excess IGF2 during development [
15,
16]. IGFBPs regulate the biological activity of IGFs by controlling IGF efflux from the vascular space, regulating IGF clearance, and modulating interaction of IGFs with their receptors [
17,
18]. The liver is a major site of synthesis for IGFBP1. IGFBP1 has been proposed as a dynamic regulator of IGF bioactivity
in vivo. IGFBP1 may be partly responsible for the decrease in lean body mass observed in catabolic/inflammatory conditions associated with low IGF1. Proinflammatory cytokines can directly enhance IGFBP-1 synthesis [
19,
20]. IGFBP2 also modulates IGF1 activity [
21]. In the brain, IGFBPs are produced by neurons, glial cells, choroid plexus epithelia, and cerebral blood vessels [
3]. IGFBP2 is the most abundant IGFBP in the CNS/cerebrospinal fluid (CSF) [
18].
There is evidence that reduced IGF1 levels correlate with impaired cognitive function in adults [
22,
23]. Alzheimer’s patients show increased serum TNFα and decreased IGF1 levels, with the two showing a negative correlation [
22]. In a study of motor neuron disease, CSF (but not serum) levels of IGF1 were diminished [
23]. Uptake of circulating IGF1 into the CSF can be substantial and circulating IGF1 could aid brain function [
24,
25]. This may explain why circulating IGF1 levels correlate with deterioration of cognitive functioning in elderly patients [
26].
IGFs and related factors have been studied in the context of HIV infection and associated diseases, as they may play a role in the immune response to HIV and influence the rate of HIV disease progression [
27,
28]. HIV/AIDS is associated with reduced IGF1 and IGF2 levels [
7,
28]. Dysregulation of IGF1 also contributes to metabolic syndromes and lipodystrophy in these individuals [
7,
29,
30]. IGFBP1 and IGFBP2 may also function as proinflammatory factors correlating with HIV disease activity [
31-
33]. The effect of antiretroviral therapy (ART) on IGF proteins can also be substantial [
34,
35]. Specifically, certain types of ART have been linked to lipodystrophy [
36,
37], vitamin D deficiency, and reduced serum IGF1 levels [
38,
39].
Little or no information is available on the role of IGFs and related proteins in neuroAIDS. HIV-associated neurocognitive disorder (HAND) affects approximately 50% of HIV+ individuals through several different mechanisms [
40-
42]. Intact virus and viral components as well as host inflammatory mediators and ART have all been implicated. There is a pressing need to identify clinically useful biomarkers for HAND. Studies by this and other laboratories suggest that depletion of neuronal growth factors might contribute to the development of HAND. Specifically, we have shown that the essential CNS growth factor progranulin is deficient in the CSF of HIV-undetectable adults with neurocognitive impairment (NCI) but not HIV-detectable adults with HAND [
43,
44]. Both progranulin and IGF1 are produced by microglia and modulated by inflammatory mechanisms [
13,
45]. A related protein IGF2 can replace IGF1 for IGF1R signaling and neuroprotection [
13]. IGF2R that is normally expressed by neurons alone in the CNS is highly upregulated in microglia in HIV encephalitis and functions as a cofactor for HIV infection [
46]. High levels of (soluble) IGF2R are present in blood, urine, and amniotic fluid [
47,
48], suggesting its potential usefulness as a biomarker.
The purpose of this study was to determine whether IGF1 or IGF2 deficiency exists in HIV+ individuals, especially in the context of neurocognitive dysfunction (HAND) and to examine the relationship between IGFs (IGF1 and IGF2) and IGFBPs (IGFBP1 and IGFBP2) in a well-characterized HIV+ cohort. We hypothesized that reduced levels of IGF1 in plasma or CSF (or both) would be associated with HAND and that chronic inflammation state in HIV infection would contribute to this. In addition, given the novel finding that IGF2R is upregulated in activated microglia and macrophages and functions as an HIV cofactor, we examined the association between IGF2R levels and HIV viral load (VL).
Methods
CHARTER subjects
CSF and plasma pairs were collected from 107 adults enrolled in the cross-sectional component of the CNS HIV Antiretroviral Therapy Effects Research (CHARTER) cohort, a six-center, US-based project funded by the National Institutes of Health. The project was approved by the institutional review board (IRB) of each participating institution (Johns Hopkins University, Mt. Sinai School of Medicine, University of California, San Diego, University of Texas Medical Branch-Galveston, University of Washington, and Washington University), as well as Albert Einstein College of Medicine. All participants provided written informed consent. Individuals with abnormal serum glucose, HCV seropositivity, or current substance use (based on urine drug screening) were excluded. Individuals receiving human growth hormone (GH), GH releasing factor, or hematopoietic growth factors such as erythropoietin were also excluded.
Neuropsychological testing
All participants completed a comprehensive neurocognitive test battery (speed of Information processing, learning and memory, executive functions, language, working memory, motor, self-report questionnaires, psychiatric interviews), covering seven cognitive domains that are commonly affected by HIV. The best available normative standards were used, which correct for effects of age, education, gender, and ethnicity, as appropriate. Test scores were automatically converted to demographically corrected standard scores (T-scores) using available computer programs. To classify presence and severity of neurocognitive impairment, a published objective algorithm that has been shown to yield excellent interrater reliability in previous multisite studies [
49] was applied. This algorithm conforms to the Frascati criteria for diagnosing HAND [
42], which requires presence of a least mild impairment in at least two domains.
IGF1 ELISA
IGF1 levels were quantified by enzyme-linked immunosorbent assay (ELISA, ALPCO Diagnostics, Salem, NH, USA: cat # 22-IGFHU-E01, sensitivity 90 pg/ml) using the core service of the Einstein Diabetes Research Center (Biomarker Analytic Research Core). We measured total IGF1 levels, that is, free and IGFBP-associated, following the manufacturer’s instructions. Briefly, ALPCO kits separate IGF1 from IGFBPs by acidification. Re-association of IGF1 with IGFBPs (after restoration of neutral pH) is prevented with an excess of IGF2. IGF2 blocks the access of IGF1 to the IGFBP binding sites without interfering with the IGF1-specific antibodies used in the assay. Samples were assayed in duplicate.
IGF2 ELISA
IGF2 levels were also quantified by ELISA (ALPCO Diagnostics, NH: cat # 22-IGF-E30, sensitivity 20 pg/ml) using the core service of the Einstein Diabetes Research Center (Biomarker Analytic Research Core, Bronx, NY, USA). Similar to IGF1 ELISA, results do not depend on the binding protein concentration of the sample. There is no cross-reactivity with IGF1. IGF2 was separated from the binding proteins by acidification, and re-association of the freed binding proteins with IGF2 was prevented by excess IGF1. Samples were assayed in duplicate.
IGF2R ELISA
Soluble IGF2R levels were quantified by ELISA (R&D Systems, Minneapolis, MN, USA: cat #DY 2447, sensitivity approximately 150 pg/ml), following the manufacturer’s instructions. The samples were diluted until the values fell within the linear range of detection (approximately 1:80 for plasma and approximately 1:10 for CSF).
Other biomarker quantifications
Concentrations of inflammatory mediators were determined in the plasma and CSF samples of all 107 subjects, as previously reported [
43]. A multiplex bead array quantified concentrations of IFNγ, IL-1β, IL-4, IL-6, IL-10, IL-13, IL-17, TNFα, CCL2/MCP-1, CXCL10/IP-10, IGFBP1, and IGFBP2 (EMD Millipore, Billerica, MA, USA). Samples were diluted until the final concentrations were within the linear range of detection for the assay. Progranulin was measured by ELISA (DY2420, R&D Systems), as reported [
43]. All samples were tested in duplicate and concentrations interpolated from a standard curve constructed by four-parameter fitting of internal standards.
Data analysis
Inflammatory mediators whose levels were below the detection limit of the assay in >90% of samples were excluded from analysis. These were IL-1β, IL-4, and IL-13 for plasma and IL-1β, IL-4, IL-13, IFNγ, and IL-17 for CSF. Undetectable HIV VL designates HIV RNA levels below 50 copies/ml, the lower limit of quantification (Amplicor, Roche Diagnostics, Indianapolis, IN, USA).
For normalization, plasma IGFBP1, plasma IGF2R, CSF IGF2, and CSF IGFBP2 concentration values were transformed with natural logarithms, and CSF IGF2R values were square-root transformed. Differences in demographic characteristics, AIDS status, current CD4 count, HIV VL, alcohol dependence, albumin, AST, ALT, and IGF1 levels in the plasma and CSF between HIV-infected individuals with and without neurocognitive impairment were examined using t-tests or Wilcoxon rank-sum tests for continuous variables and using chi-squared or Fisher’s exact tests for categorical variables. Initial analysis of associations of demographic and disease characteristics with plasma and CSF IGF protein concentrations were examined using t-tests or Wilcoxon rank-sum tests for categorical variables and Spearman correlation coefficients for continuous variables. Linear regression models or analysis of covariances (ANCOVA) were used to examine associations of IGF proteins with neurocognitive status adjusted for demographics, HIV VL, and inflammatory mediators. Variables with a P-value ≤0.2 in initial analysis were included in multivariate analysis. Because neither log-transformation nor square root transformation normalized CSF IGF1, CSF IGFBP1, or plasma IGFBP2, non-parametric analysis of covariances was used for these three variables. Statistical analyses were performed using SAS (version 9.1, SAS Institute, Cary, NC, USA) software.
Discussion
Our analyses of IGF proteins included association with demographic and clinical parameters which confirmed many of the previously known associations, such as with age, gender, race, BMI, and liver function. As these factors have rarely been examined in the CSF, our study also provides useful information in comparison to their plasma levels. For example, while plasma IGF1 levels decreased with aging (as reported [
26]), CSF IGF2R (as well as IGF1 and IGF2 levels) tended to
increase with aging. Black race was associated with lower plasma levels of IGF family proteins (as reported [
50,
51]), and their levels (IGF1, IGF2, and IGF2R) measured in the CSF were also lower in Blacks. Lower IGFBP1 levels in both plasma and CSF were strongly (
P < 0.001) associated with high BMI [
52]. Alternations of many CSF IGF family proteins were also associated with alcohol dependence. While we cannot explain exact mechanisms for these associations, our data suggest that IGF proteins play dynamic and complex roles under physiologic and pathologic conditions and form the basis for future studies.
There are well-known associations between certain IGF family proteins, suggesting that their productions are coordinately regulated. Our study confirms some of these associations and provides additional information on their relationships (see the ‘
Results’ section including Additional file
1). For example, in plasma, IGF1 and IGF2 were closely correlated and IGFBP1 and IGFBP2 were closely correlated (Additional file
1: Table S1B). Inflammatory mediators showed positive correlations with IGFBP1/IGFBP2 and negative correlations with IGF1/IGF2, indicating that inflammation suppresses IGF1/IGF2 production (Additional file
1: Table S1B). The inflammatory mediators that showed correlations included IL-6, which is an established biomarker for systemic inflammation in HIV+ individuals [
53], as well as the newly defined proinflammatory cytokine IL-17. These findings together suggest that inflammation is a highly significant underlying mechanism that regulates plasma IGF protein levels. IGF2R, which has not been examined in the context of HIV until now, showed a correlation with IGFBP1 and IGFBP2, as well as all of the inflammatory mediators examined (IL-17, IFNγ, IL-10, TNF-α, IL-6, and MCP-1) except IP-10 on univariate analyses (Additional file
1: Table S1B). These results suggest that plasma IGF2R is induced by inflammatory mediators and is a potential biomarker for systemic inflammation in HIV+ individuals.
In our study, we also find that ART use was associated with alternations of plasma and CSF IGF protein levels. Specifically, higher plasma IGF2, CSF IGF2, and CSF IGF2R levels were associated with ART use on univariate analysis (Additional file
1: Table S1A and Table
4), possibly reflecting virologic and immunologic improvements associated with ART. Another important change in the IGF axis relates to the types of ART. For example, in our cohort, lower IGF1 levels were strongly associated with non-nucleotide reverse transcriptase inhibitors (NNRTI)-based regimen compared to protease inhibitors (PI)-based regimen (
P = 0.008). This is consistent with the notion that efavirenz-induced vitamin D deficiency caused circulating IGF-1 deficiency in our cohort, as (1) efavirenz has been reported to induce vitamin D deficiency [
39], (2) vitamin D is a known modulator of circulating IGF1 [
38], and (3) efavirenz is the main NNRTI used by CHARTER subjects. These results add significantly to the notion that toxicity related to ART can contribute to neurocognitive impairment, in part, through growth factor depletion.
Analyses of CSF revealed that IGF2R levels correlated with IGF2 but not with other proteins or with inflammatory mediators (Additional file
1: Table S2B). These results may reflect that the source of CSF IGF2R in this cohort is mostly neuronal with little or no contributions from microglia or macrophages. This is consistent with our previous study which showed that
de novo IGF2R induction in microglia is found in HIV encephalitis but not in most HIV+ (but HIVE-) brains [
46]. In the CSF, IGF2 appears to play a more significant role than IGF1. First, the IGF2: IGF1 ratios were considerably higher in the CSF (approximately 40:1) than plasma (approximately 3:1), indicating intrathecal production. Second, CSF IGF2 showed correlations with most other CSF factors including IGF2R, IGF1, IGFBP1, IGFBP2, IL-10, IP-10, and TNFα (Additional file
1: Table S2B). Furthermore, CSF IGFBP1 correlated with CSF IGFBP2, and both IGFBPs correlated with inflammatory mediators. The correlations were most significant with IGFBP2, whose CSF levels were higher than those in plasma, again indicating intrathecal production. The emerging picture is then IGF2 and IGFBP2 are the major IGF family proteins in the CSF/CNS. The source of IGF2 in the CSF compartment includes the choroid plexus and the meninges [
14]. We found
in vitro that while both IGFs were equally neuroprotective, IGF2 was
upregulated and IGF1 was
downregulated by proinflammatory mediators in cultured microglia [
13], suggesting that the two IGFs may not (necessarily) be co-regulated in the CNS. The potential source of IGF1 in the brain and CSF compartment is also multiple including monocyte-lineage cells and other mesenchymal cells [
13]. However, because intrathecal IGF1 production is minimal compared to systemic circulation, the most important source of CSF IGF1 might be circulating IGF1 [
18]. Furthermore, as IGF2R is a scavenging receptor for both IGFs, upregulated IGF2R expression in CSF/CNS (in HIV encephalitis, for instance [
46]) could function as a significant modulator of IGF protein expression. For instance, soluble IGF2R shed from cells (such as activated microglia and macrophages) can capture IGF2 and IGF1, effectively blocking their binding to cognate signaling receptor IGF1R. In this regard, the soluble IGF2R we detect in plasma/CSF should be considered a kind of IGF1R antagonist.
Our study confirms some of the reported IGF axis alterations associated with HIV infection and also adds new data. The association of low plasma IGF1 (or IGF2) and high IGFBPs with HIV progression (AIDS, low CD4 T cell counts) have been well established [
28,
31-
33,
54-
56]. We also show for the first time that increased plasma IGF2R is associated with detectable plasma VL and low CD4+ T cell counts. As
de novo IGF2R expression occurs in human microglia in HIV encephalitis (a model of tissue macrophages HIV infection) and IGF2R is a cellular cofactor for HIV replication [
46], our biomarker data supports that IGF2R (likely induced by inflammatory mediators) boosts HIV replication (hence detectable VL) and induces CD4+ T cell depletion. Together, these results show a novel finding that IGF2R is an important new player in the IGF axis that connects inflammation, HIV replication, and growth factor depletion.
We initially hypothesized that lower CSF IGF1 might be associated with NC impairment in HIV+ individuals but our hypothesis has not been borne out by our analyses. There are several possible reasons for this. First, the CSF IGF1 levels are very low (approximately 1/100 to 1/200 of plasma levels) and therefore proving our hypothesis may require a much larger cohort. Second, the IGF axis has been shown to exhibit great inter-individual heterogeneity. For example, longitudinal analysis of 1,422 HIV+ women showed only a trend for association between low IGF1 levels and HIV disease progression [
57]. In our analysis of 107 HIV+ subjects, plasma IGF1 and IGF2 levels as well as CSF IGF1 levels were lower in NC impaired (
n = 58) compared to those in NC unimpaired (
n = 49) subjects, though they were not statistically significant. Of the three, plasma IGF1 showed the largest difference between the two groups (186.5 ± 63.28
vs. 200.1 ± 43.7,
P = 0.127), possibly suggesting that a much larger cohort may have shown a significant association between low plasma IGF1 and NC impairment. Furthermore, a combined growth factor deficiency (rather than a single deficiency) may contribute to significant functional outcome. In this regard, positive correlation between CSF IGF1 and CSF progranulin in our cohort (
r = 0.26,
P = 0.008) (Additional file
1: Table S2B) suggests a common mechanism of growth factor regulation in HIV+ individuals and lend further support to this hypothesis. Given the number of studies hinting at the role of CNS growth factor deficiencies in HAND [
43,
58,
59], further studies are warranted to examine their potential roles in the pathogenesis and this complex CNS disorder.
Acknowledgements
This study was supported by the NIH grants K01MH084705, R01MH55477, Einstein CFAR (P30AI051519), and the CNS HIV Antiretroviral Therapy Effects Research (CHARTER). We are grateful to Robin Squeglia at the Einstein Biomarker Analytic Research Core (BARC) Laboratory for assisting with the IGF assays and Dr. Susan Morgello for her support throughout the study. CHARTER was supported by the NIH grants N01 MH22005, HHSN271201000036C, and HHSN271201000030C. The views expressed in this article are those of the authors and do not reflect the official policy or position of the United States Government.
The CNS HIV Antiretroviral Therapy Effects Research (CHARTER) group is affiliated with Johns Hopkins University; the Icahn School of Medicine at Mount Sinai; University of California, San Diego; University of Texas, Galveston; University of Washington, Seattle; Washington University, St. Louis; and is headquartered at the University of California, San Diego and includes: Director: Igor Grant, M.D.; Co-Directors: Scott L. Letendre, M.D., Ronald J. Ellis, M.D., Ph.D., Thomas D. Marcotte, Ph.D.; Center Manager: Donald Franklin, Jr.; Neuromedical Component: Ronald J. Ellis, M.D., Ph.D. (P.I.), J. Allen McCutchan, M.D.; Laboratory and Virology Component: Scott Letendre, M.D. (Co-P.I.), Davey M. Smith, M.D. (Co-P.I.).; Neurobehavioral Component: Robert K. Heaton, Ph.D. (P.I.), J. Hampton Atkinson, M.D., Matthew Dawson; Imaging Component: Christine Fennema-Notestine, Ph.D. (P.I.), Michael J Taylor, Ph.D., Rebecca Theilmann, Ph.D.; Data Management Component: Anthony C. Gamst, Ph.D. (P.I.), Clint Cushman; Statistics Component: Ian Abramson, Ph.D. (P.I.), Florin Vaida, Ph.D., Reena Deutsch, Ph.D.; Johns Hopkins University Site: Justin McArthur (P.I.), Vincent Rogalski; Icahn School of Medicine at Mount Sinai Site: Susan Morgello, M.D. (Co-P.I.) and David Simpson, M.D. (Co-P.I.), Letty Mintz, N.P.; University of California, San Diego Site: J. Allen McCutchan, M.D. (P.I.), Kaori Phillips, B.S.N.; University of Washington, Seattle Site: Ann Collier, M.D. (Co-P.I.) and Christina Marra, M.D. (Co-P.I.), Trudy Jones, M.N., A.R.N.P.; University of Texas, Galveston Site: Benjamin Gelman, M.D., Ph. D. (P.I.), Eleanor Head, R.N., B.S.N.; and Washington University, St. Louis Site: David Clifford, M.D. (P.I.), Muhammad Al-Lozi, M.D., Mengesha Teshome, M.D.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HS, YL, and SCL designed the study and interpreted data. HS and NC performed ELISA. YL and HS performed statistical analyses. HS, YL, SL, and SCL wrote the paper. All authors read and approved the final manuscript.