Association of serum hepcidin with prostate-specific antigen levels in men from high Andean cities of Peru

Introduction

People living at high altitude (HA) show different mechanisms to adapt to hypoxic environments to compensate for the decrease in partial pressure of oxygen.[1,2] In this acclimatization process, the activation of the endocrine system plays a crucial role.[3] For instance, the sex hormone testosterone is activated in a physiological response to HA exposure regulating hemoglobin (Hb) levels to stimulate erythropoiesis.[4] However, in maladapted individuals, it may cause excessive erythrocytosis (EE) (Hb ≥ 21 g/dL), a solid indicator of chronic mountain sickness (CMS).[5]

Nevertheless, the liver hormone hepcidin controls iron homeostasis and contributes to HA adaptation by sustaining iron availability, which is critical for Hb synthesis during erythropoiesis.[6]

Hepcidin reduces the release of intracellular free iron from the reticuloendothelial system and inhibits intestinal iron absorption without affecting iron reserve. This mechanism acts by blocking the iron transporter ferroportin.[7,8] However, hepcidin synthesis is not restricted only to the liver since its expression was too reported in the prostate epithelial cells in an autocrine regulation.[9,10] Moreover, in patients with prostate cancer, hepcidin expression contributes not only to tumor-associated anemia but also to tumor growth, angiogenesis, and metastasis.[11] This process is mediated by inhibiting ferroportin and increasing intracellular iron retention.[6,12,13] Thence, the proposal of hepcidin as a new prognostic factor for prostate cancer.[11,14]

On the other hand, a biomarker currently used for diagnosing prostate cancer is the prostate-specific antigen (PSA). This biomarker has been slightly studied in HA residents,[15] and it is unknown how it could predict prostate disease in highlands. Scientific literature reported that in healthy HA landers exposed to chronic hypobaric hypoxia, the levels of PSA were lower than sea level landers.[16] Also, in miners with normal values of PSA (≤4 ng/mL) exposed to intermittent hypobaric hypoxia, the PSA levels were lower when the altitude difference between their works and their camps was extensive.[17] Notwithstanding, the accuracy of PSA in predicting prostate illness in men exposed to hypobaric hypoxia is unclear, and complementary biomarkers might be required.

This article analyzed how hepcidin is associated with PSA in healthy men at three high-altitude cities and one sea-level city to test whether hepcidin could be a complementary biomarker in the diagnosis of prostate disease. We hypothesized that hepcidin levels correlate with PSA at low and HAs.

Materials and Methods

Results

The characteristics of the participants are shown in Table 1. The mean age was 37 years, with a higher proportion of obese men in Huancayo city (38.5%). Hb increased with altitude and were higher in Huancayo (19.4 g/dL), Puno (17.6 g/dL), and Cerro de Pasco (18.6 g/dL) than Lima (15.3 g/dL). However, the Hb levels in Puno (3800 m) were lower than in Huancayo (2380 m) and Cerro de Pasco (4320 m). Only Hb in Huancayo and Cerro de Pasco was not statistically different (p ≤ 0.05).

Table 1 Participant characteristics

The proportion of polycythemia (Hb 17–20 g/dL) was 61.5%, 31.8%, and 43.5% for Huancayo, Puno, and Cerro de Pasco, respectively. The proportion of EE (≥21 g/dL) was 7.7%, 4.5%, and 8.7% for Huancayo, Puno, and Cerro de Pasco, respectively. All high-altitude cities had a CMS score of 2 (Puno) or 3 (Huancayo and Cerro de Pasco), while Lima had 0.

The testosterone levels dropped from 6.5 ng/mL in Lima (low altitude) to 3.6 ng/mL in Huancayo, 2.2 ng/mL in Puno but 5.3 ng/mL in Cerro de Pasco without a tendency to altitude. Although significant only at P = 0.133, the hepcidin levels in Huancayo (23.4 ng/mL) and Cerro de Pasco (18.6 ng/mL) tended to be higher than in Puno (16.3 ng/mL) and in Lima (15.6 ng/mL).

The PSA levels were higher at sea-level and dropped from 0.4 ng/mL in Lima to 0.3 ng/mL in Puno and 0.2 ng/mL in Huancayo and Cerro de Pasco (P < 0.05). Hepcidin levels did not differ statistically among cities a different altitude, although they tended to be higher at HA.

Bivariate association analysis of hepcidin and PSA by city and age [Figure 1] showed no significant correlation between hepcidin concentration and PSA levels at different altitudes. However, hepcidin positively correlated with Hb, CMS score, and BMI (p ≤ 0.05) but not with testosterone or partial pressure of oxygen (SpO₂) (p > 0.05) [Figure 2].

Close

Figure 1

(a and b) Association between hepcidin and PSA by categorized age and city. City: 1: Lima (<150 m), 2: Huancayo (2380 m), 3: Puno (3800 m), 4: Cerro de Pasco (4320 m). Categorized age: <40 years; ≥40 years. Rho (ρ): Spearman coefficient. PSA: Prostate-specific antigen. Statistical significance: P ≤ 0.05

Export to PPT

Close

Figure 2

Correlation between hepcidin and biological parameters by age. City: 1: Lima (<150 m), 2: Huancayo (2380 m), 3: Puno (3800 m), 4: Cerro de Pasco (4320 m). (a) Hepcidin versus Age, (b) Hepcidin versus Testosterone, (c) Hepcidin versus Hemoglobin, (d) Hepcidin versus CMS-score, (e) Hepcidin versus SpO₂, (f) Hepcidin versus BMI. Hb: Hemoglobin; CMS: Chronic mountain sickness; SpO₂: Pulse oxygen saturation, BMI: Body mass index. Rho (ρ): Spearman coefficient. Statistical significance: P ≤ 0.05

Export to PPT

Furthermore, we performed a linear mixed-effect analysis [Table 2]. There were five missing observations for hepcidin (n = 65). The null model showed the intraclass correlation coefficient was 0.03, which indicates that only 3% of the total variance in hepcidin was due to the variability of each city. We found no association between hepcidin and PSA, even adjusted by covariates (age, BMI, Hb, and SpO₂) (Models 1–4). Even though, BMI and Hb are positively correlated with hepcidin (Model 3 and 4), these two variables did not contribute to the association between hepcidin and PSA.

Table 2 Multivariable linear mixed models of hepcidin with random intercept effects of cities

In this analysis, the linear mixed-effect models did not include testosterone due to collinearity with hemoglobin.

Discussion

This pioneer investigation analyzes the relationship between hepcidin and PSA in high-altitude landers. Our study explored this association based on the fact that prostate malignancy is related to oxygen scarcity in the tissue microenvironment due to low oxygen availability and increased oxygen consumption resulting from rapid cell proliferation rates.[22,23]

Another fact that supported our hypothesis is that hepcidin can be secreted by prostate cancer cells.[12,24] Synthesis and secretion of hepcidin from prostate epithelial cells are markedly increased in this organ cancer status.[6] In addition, it has been reported that when troops are rapidly deployed to HA, energy consumption is affected and an equivalent activity at HA requires greater effort and more energy than it does at sea level.[25] Likewise, higher levels of serum and urine hepcidin are presented in athletes after prolonged periods of training,[26] especially during the first 3 h post exercise under hypoxic conditions.[27] Nevertheless, acclimatization to HA also involves a negative energy balance that suppresses appetite and reduces energy intake.[28]

Lifelong exposure to altitude is accompanied by a lower risk for obesity, whereas altitude sojourns are generally associated with a loss of body mass.[29] However, in our study, the higher number of obesity men was at HA cities.

We expected a negative correlation between hepcidin and PSA, not only by hypoxia but also by obesity (which is a BMI category), because it is negatively correlated with PSA.[30] However, we did not find a statistical association between hepcidin and PSA, or a trend with altitude. This probably due to the age of the participants, who were relatively young population, and their age distribution did not show a positive correlation with hepcidin; even the analysis by altitude city.

Obesity has the critical ability to predict stress-associated diseases through triglycerides and cholesterol biomarkers.[31] In this regard, obesity can be considered as a low grade inflammatory state, which stimulates the production of inflammatory markers such as CRP that can up-regulate hepcidin synthesis[32] and related with prostatic diseases,[33] which unfortunately were not analyzed in this study.

Considering that, in this study, there was a significant prevalence of obesity in the high cities (Huancayo, Puno and Cerro de Pasco), we analyzed the relationship between hepcidin and BMI and found a positive association. However, we did not observe a significant correlation between hepcidin and PSA by BMI categories, nor significant correlation between PSA and BMI. Although, we did observe a higher concentration of PSA at sea level (without obese cases) compared to the higher altitude cities, contrarily to the known negative association between PSA and BMI.[34,35]

Because extended sojourns to HA, results in increased erythropoiesis and an increased physiological requirement for iron,[28] we also investigated whether other biomarkers influence the association between hepcidin and PSA. In this sense, the multivariable linear mixed model showed that altitude accounts for only 3% of the variability of hepcidin. Thus, erythropoietin levels, ferritin, iron availability, inflammation markers,[36] and external factors not evaluated in this study might influence our results. Concerning to inflammation, the increased levels of hepcidin in response to exercise may be induced by stress and inflammatory signal.[37]

Indeed, like others before,[38] we show a positive association between hepcidin and Hb after adjusting for BMI, age, and SpO₂ as well high and significant correlation with CMS score. This observation is surprising because elevated erythropoiesis demands more iron than usual, where hepcidin diminishes, and the erythroblast-derived hormone erythroferrone suppresses hepcidin expression to meet the elevated iron demand.[36] In addition, increased Hb levels induce higher hepcidin levels, since no iron is required for erythropoiesis.[5,39] However, in CMS patients, basal hepcidin levels were similar to those observed at low and HAs.[8]

For example, the Ethiopian population has steady-state hepcidin levels that do not differ between low and HA, suggesting that natives at HA are not iron deficient.[38] The erythropoietic activity in these people seems to be relatively stable, although elevated. The erythroid regulator signals an imbalance between erythropoietic drive and iron supply rather than absolute marrow activity.[8]

Moreover, in HA the dietary iron and iron recycling from macrophages cannot match the increased requirements for iron.[7] Likewise, in acute hypoxia, these results make us hypothesize that chronic hypoxia might induce a rapid mobilization of iron from storage cells to plasma to ensure enough iron for the mounting red blood cell hemoglobinization.[40] Another hypothesis is that plasmatic hepcidin would not suppress iron-replete individuals chronically adapted to HAs.[38]

In this sense, a man well-adapted to HA should have hepcidin levels as normal as those at sea-level. Therefore, an elevated PSA concentration or a high prevalence of prostate cancer would not be expected. Unfortunately, the scientific evidence on prostate diseases at HA is still insufficient and our results did not clarify this hypothesis.

Further, it was also surprising to find a positive association of hepcidin with CMS score, and with Hb. In this regard, a manifestation of EE (Hb >21g/dL),[41] a sign of CMS, has been associated with higher levels of testosterone.[42]

Testosterone, which also plays an essential role in prostate growth and tumor hypoxia,[43,44] could suppress hepcidin and consequently increase hematocrit in a manner analogous to testosterone therapy,[45] and independently of its effects on erythropoietin or hypoxia-sensing mechanisms.[46] Alternatively, testosterone could increase iron bioavailability and Hb up to EE,[5] and it could also activate ferroportin, transferrin, and transferrin receptor (TR), improving iron bioavailability for erythropoiesis.[47]

Despite these facts, in our results, testosterone did not impact hepcidin, Hb, or CMS as did hepcidin with Hb and CMS score.

Disturbances in hepcidin regulation contribute to the pathogenesis of iron disorders in two ways: One by deficiency, which causes iron overload in hereditary hemochromatosis and non-transfused β-thalassemia; and the other one by overproduction, which is associated with iron-restricted anemia’s seen in patients with chronic kidney disease, chronic inflammatory diseases, and some cancers[48] and several agents aimed at manipulating this pathway have been designed, with some already in clinical trials.[49] However, these pathologies have only been evaluated based on hepatic synthesis, the main determinant of plasma hepcidin concentrations.

Furthermore, regarding molecular studies in prostate cells that showed higher hepcidin levels,[12] their approach allows no differentiation between prostatic and hepatic hepcidin secretion.

Murine studies suggest that hepcidin mRNA levels in the pancreas are 1000 times lower than in the liver.[50] Considering the organ size of the human liver (ca. 1.5 kg) and pancreas (ca. 0.1 kg)[51] and the total number of hepcidin-producing cells, we predict that serum hepcidin levels obtained in our study are predominantly liver-derived. Thus, we cannot affirm that the analyzed serum hepcidin samples come only from the prostate. In addition, prostate hepcidin has an autocrine function,[6] and all its mechanisms developed in prostate cancer cells, such as decreased cell surface ferroportin and increased intracellular iron retention, would only act at the cellular level within the prostate. However, it is not yet known whether differences in total hepcidin and alterations in Hb could influence prostate metabolic pathways.

The limitations in our study are related to the sample size, the lack of measurements of erythropoietin, ferritin, iron availability, and inflammation markers. The time to withdraw the serum testosterone samples (which is hormone time-dependent), and the fact that we have tested this association in healthy men but not in prostate cancer (PCa) patients, are additional limitations.

Even though one research reported higher hepcidin levels correlating with higher PSA levels in prostate patients,[11] we support the hypothesis that hepcidin concentration may depend on the personal and physiologic characteristics to adapt to hypobaric hypoxia. Moreover, we cannot exclude the possibility that this condition would be worst in PCa patients who reside at HA. Further research is needed to demonstrate whether hepcidin would be a good candidate as an adjunctive biomarker in the diagnosis of prostatic diseases in HA residents.

Conclusion

Summing up the aforementioned limitations with our findings and additional analysis (Appendix), it would be difficult to affirm that hepcidin would be a promising biomarker for prostatic illness at HA.

Authors’ Declaration Statements