Background

Prostein (P501S), also termed solute carrier family 45 member 3 (SLC45A3) is a protein composed of 553 amino acids which is coded by the SLC45A3 gene at chromosome 1q32.1 [1]. Its function is not well known but some data suggest a role in transmembrane transport of sugars [2]. Prostein is predominantly expressed in the prostate, where its expression is androgen regulated [3]. Prostein is the second most common 5′ partner gene in ETS Transcription Factor ERG (ERG) rearrangements in prostate cancer after Transmembrane Serine Protease 2 (TMPRSS2) [4, 5], another constitutively expressed androgen regulated gene in prostate epithelium [6]. In the brain, prostein plays a role in regulating the lipid metabolism of oligodendrocytes and myelin [7].

A high level of prostein expression is a common feature in prostate cancer. Amanda et al. [8] described prostein positivity in 97% of 59 analyzed prostate cancers. Queisser et al. [9] found prostein expression in 96% of 79 prostate cancers. Sheridan et al. [10] reported prostein positivity in 99% of 53 metastatic prostatic carcinomas. Based on these data, prostein immunohistochemistry (IHC) has been suggested as a diagnostic tool for the distinction of prostatic adenocarcinoma from other tumors. This notion is also supported by data describing high specificity of prostein expression for prostate cancer. For example, Garudadri et al. [11] described a 100% specificity of prostein IHC in a study on 100 prostatic carcinomas and 60 normal and cancerous extra-prostatic tissues. In an analysis of 600 tumors from 20 sites of origin, Mochizuki et al. [12] found prostein positivity in 30 of 30 prostate adenocarcinomas but in only one tumor each of 30 hepatocellular carcinomas and of 30 invasive breast cancers of no special type (NST). Kalos et al. [3] did not detect prostein staining in 3,454 samples of more than 130 tumor entities and subentities while 94% of 60 analyzed prostate cancers showed prostein positivity. Osunkoya et al. [13] did not find prostein positivity in any of 9 colorectal adenocarcinomas infiltrating the prostate. Srinivasan et al. [14] did not see any prostein positivity in 132 urothelial carcinomas. However, Arnesen et al. [15] found prostein positivity in 11 of 14 Sertoli-Leydig or Leydig cell tumors of the testis and ovary and Chuang et al. [16] reported prostein positivity in 7 of 41 invasive urothelial carcinomas.

To further corroborate the potential diagnostic utility of prostein IHC, a comprehensive survey of prostein immunostaining in an even broader range of tumor types is desirable. We therefore evaluated prostein expression in more than 19,000 tumor tissue samples from 152 different tumor types and subtypes as well as 76 different non-neoplastic tissue types by IHC in a tissue microarray (TMA) format.

Materials and methods

Tissue microarrays (TMAs)

Our normal tissue TMA was composed of 8 samples from 8 different donors for each of 76 different normal tissue types (608 samples on one slide). The cancer TMAs contained a total of 19,202 primary tumors from 152 tumor types and subtypes. The composition of both normal and cancer TMAs is described in detail in the “Results” section. Clinico-pathological data including pathological tumor stage (pT), grade, lymph node status (pN), lymphatic vessel (L) and blood vessel (V) infiltration were available for 327 gastric, 2,139 breast, and 2,351 colorectal carcinomas. All samples were from the archives of the Institutes of Pathology, University Hospital of Hamburg, Germany, the Institute of Pathology, Clinical Center Osnabrueck, Germany, and Department of Pathology, Academic Hospital Fuerth, Germany. Tissues were fixed in 4% buffered formalin and then embedded in paraffin. TMA tissue spot diameter was 0.6 mm. The use of archived remnants of diagnostic tissues for manufacturing of TMAs and their analysis for research purposes as well as patient data analysis has been approved by local laws (HmbKHG, § 12) and by the local ethics committee (Ethics commission Hamburg, WF-049/09). All work has been carried out in compliance with the Helsinki Declaration.

Immunohistochemistry

Freshly cut TMA sections were immunostained on one day and in one experiment. Slides were deparaffinized with xylol, rehydrated through a graded alcohol series and exposed to heat-induced antigen retrieval for 5 min in an autoclave at 121 °C in pH 9.0 DakoTarget Retrieval Solution™ (Agilent, CA, USA; #S2367). Endogenous peroxidase activity was blocked with Dako Peroxidase Blocking Solution™ (Agilent, CA, USA; #52,023) for 10 min. Primary antibody specific for prostein (rabbit recombinant monoclonal, MSVA-460R, MS Validated Antibodies, Hamburg, Germany; #5241-460R) was applied at 37 °C for 60 min at a dilution of 1:150. For the purpose of antibody validation, the normal tissue TMA was also analyzed by the rabbit recombinant monoclonal prostein antibody EPR4795(2) (Abcam, Cambridge, UK; #ab137065) at a dilution of 1:150 and an otherwise identical protocol. Bound antibody was then visualized using the EnVision Kit™ (Agilent, CA, USA; #K5007) according to the manufacturer’s directions. The sections were counterstained with haemalaun. For normal tissues, the staining intensity of positive cells was semi-quantitively recorded (+, ++, +++). For tumor tissues, the percentage of prostein positive neoplastic cells was estimated, and the staining intensity was semi-quantitatively recorded (0, 1+, 2+, 3+). For statistical analyses, the staining results were categorized into four groups. Tumors without any staining were considered negative. Tumors with 1 + staining intensity in ≤ 70% of tumor cells or 2 + intensity in ≤ 30% of tumor cells were considered weakly positive. Tumors with 1 + staining intensity in > 70% of tumor cells, 2 + intensity in 31-70%, or 3 + intensity in ≤ 30% of tumor cells were considered moderately positive. Tumors with 2 + intensity in > 70% or 3 + intensity in > 30% of tumor cells werde considered strongly positive.

Statistics

Statistical calculations were performed with JMP 16 software (SAS Institute Inc., NC, USA). Contingency tables and the chi²-test were performed to search for associations between prostein immunostaining and tumor phenotype.

Results

Technical issues

A total of 17,146 (89.3%) of 19,202 tumor samples were interpretable in our TMA analysis. Non-interpretable samples demonstrated lack of unequivocal tumor cells or loss of the tissue spot during technical procedures. A sufficient number of samples (≥ 4) of each normal tissue type was evaluable.

Prostein in normal tissues

Prostein staining was always granular, cytoplasmic and predominantly perinuclear (“endoplasmatic reticulum pattern”). The staining was particularly strong in acinar cells of the prostate and occurred at lesser intensity in surface epithelial cells of the stomach, in goblet cells of the respiratory epithelium of the lung and (weaker) in bronchial glands, as well as in a subset of epithelial cells of the adenohypophysis. A weak prostein staining was also seen in few colorectal epithelial cells (not in all samples) and in a subset of pancreatic islet cells. A perinuclear granular cytoplasmic prostein positivity also occurred in a small fraction of (monocytic) cells in the spleen and in few cells of lymph nodes. In the brain, some glia cells showed a perinuclear granular cytoplasmic prostein staining. Representative images are shown in Fig. 1. All these findings were seen by both antibodies, MSVA-460R and EPR4795(2). An additional cytoplasmic staining in the placenta and in testicular cells of the spermatogenesis was only seen by EPR4795(2) (Supplementary Fig. 1) and therefore considered an antibody-specific cross-reactivity of EPR4795(2). Prostein immunostaning was absent in skeletal muscle, heart muscle, smooth muscle, myometrium of the uterus, corpus spongiosum of the penis, ovarian stroma, fat, skin (including hair follicles and sebaceous glands), oral mucosa of the lip, surface epithelium of the oral cavity and the tonsil, transitional mucosa of the anal canal, ectocervix, squamous epithelium of the esophagus, urothelium of the renal pelvis and urinary bladder, decidua, placenta, thymus, tonsil, gall bladder, liver, parotid gland, submandibular gland, sublingual gland, duodenum, small intestine, appendix, colorectum, kidney, seminal vesicle, testis, epididymis, breast, endocervix, endometrium, fallopian tube, adrenal gland, parathyroid gland, and the neurohypophysis.

Fig. 1
figure 1

Prostein immunostaining of normal tissues. Prostein staining was always granular, cytoplasmic and predominantly perinuclear (“endoplasmatic reticulum pattern”). The panels show a particularly strong prostein staining of acinar cells of the prostate (A) while the staining is less intense in surface epithelium of the stomach (B). An even weaker prostein positivity (not always involving all samples and all cells) can also be seen in colorectal epithelium (C), pancreatic islet cells (D), epithelial cells of the adenohypophysis (E), respiratory epithelium of the lung (F), and in glia cells of the brain (G). An intense perinuclear prostein staining also occurs in a subset of monocytic cells of the spleen (H)

Full size image

Prostein in cancer tissues

Similarly, as in normal tissues, prostein immunostaining was typically cytoplasmic, granular and perinuclear in tumors. Prostein positivity, and especially a strong prostein staining was predominantly seen in prostatic adenocarcinomas. 93% of primary prostate cancers and 63% of recurrent prostate cancers showed a strong prostein immunostaining while 98% of primary prostate cancers and 94% of recurrent prostate cancers showed at least a weak positivity. Prostein staining was absent in all 18 small cell neuroendocrine carcinomas of the prostate. Prostein positivity - mostly at a lower level - was also detectable in 1,204 (7.2%) of the 16,709 analyzable extra-prostatic tumors. Of these, 922 (5.5%) showed a weak, 239 (1.4%) a moderate, and only 43 (0.3%) a strong immunostaining. Overall, 50 (34.0%) of 157 extra-prostatic tumor categories showed detectable prostein expression with 12 (8.2%) tumor categories including at least one strongly positive tumor (Table 1). Representative images of prostein positive tumors are shown in Fig. 2. Extra-prostatic tumors with highest rate of prostein positivity included different subtypes of salivary gland tumors (7.6-44.4%), neuroendocrine neoplasms (15.8-44.4%), adenocarcinomas of the gastrointestinal tract (7.3-14.8%), and biliopancreatic adenocarcinomas (3.6-38.7%), hepatocellular carcinomas (8.1%), as well as adenocarcinomas of other organs of origin (up to 21%). A graphical representation of a ranking order of prostein positive and strongly positive cancers is given in Fig. 3. A comparison between prostein expression and tumor phenotype is shown in Table 2. Detectable prostein expression was linked to high grade (p = 0.0105), HER2 positivity (p = 0.0312), and estrogen receptor negativity (p = 0.0330) in invasive breast carcinomas of no special type (NST), V0 status (p = 0.0139), right sided tumor location (p = 0.0479), and KRAS mutations (p = 0.0133) in colorectal cancer, pN0 stage (p = 0.0424) in pancreatic ductal adenocarcinoma as well as to microsatellite instability in gastric cancers (p = 0.0015).

Table 1 Prostein immunostaining in human tumors
Full size table
Fig. 2
figure 2

Prostein immunostaining in cancer. Prostein staining is usually granular, cytoplasmic and predominantly perinuclear (“endoplasmatic reticulum pattern”). The panels show a particularly strong prostein positivity in a Gleason 3 + 3 = 6 carcinoma (A) and a recurrent Gleason 5 + 5 = 10 carcinoma of the prostate (B). Prostein staining of tumor cells is less intense but still significant in samples of mucoepidermoid carcinoma of a salivary gland (C), neuroendocrine tumor of the lung (D), adenocarcinoma of the colon (E), and a muscle-invasive urothelial carcinoma of the urinary bladder (F). A distinct staining of giant cells is seen in samples of a giant cell tumor of the tendon sheet (G) and a pilomatrixoma of the skin (H)

Full size image
Fig. 3
figure 3

Ranking order of prostein immunostaining in tumors. Both the percentage of positive cases (blue dots) and the percentage of strongly positive cases (orange dots) are shown

Full size image
Table 2 Prostein and tumor phenotype
Full size table

Discussion

Our successful analysis of more than 17,000 tumors provided a comprehensive overview on the patterns of prostein expression in cancer. The predominant expression of prostein in prostate cancer was expected since studies analyzing 9-220 tumor cases had earlier identified prostein positivity in up to 100% of prostate cancers [4, 11, 17, 18]. Our positivity rate of 100% in Gleason 3 + 3 = 6, 98% in Gleason 4 + 4 = 8 and 97% in Gleason 5 + 5 = 10 prostate cancers is comparable with results from most previous studies [3, 19]. The concept that prostein IHC can be used to corroborate a suspected prostatic origin of a cancer tissue is further supported by the retained prostein expression in at least 80% of prostate cancers that recurred after hormonal therapy [19]. Sheridan et al. [10] had previously identified prostein positivity in 99% of 53 analyzed prostatic cancer metastases. Hernandez-Llodra et al. [4] have previously suggested that the few prostate cancers with reduced or absent prostein expression might harbor SLC45A3:ERG fusions and that these tumors may be characterized by poor prognosis.

The extensive analysis of non-prostatic tumors in this study identified a considerable number of tumor entities that can also express prostein. Although prostein expression was less frequent and often at markedly lower level in these tumors than in prostate cancer, the characteristic staining pattern with a distinct granular, perinuclear cytoplasmic prostein staining was always retained. The most commonly prostein positive tumors included salivary gland tumors, neuroendocrine neoplasms, various categories of gastrointestinal or biliopancreatic adenocarcinomas, hepatocellular carcinomas as well as adenocarcinomas of other organs of origin. All these tumor entities represent diagnostic options in case of a prostein positive tumor mass. It is of note that in some tumor entities, a perinuclear prostein expression was also observed in cells of monocytic origin such as for example in epitheloid cells accompanying lymphomas or in giant cells of tendon sheath tumors or in pilomatricoma. These findings fit with our observation of prostein positive monocytic cells in the spleen and the lymph node. Our data in primary and recurrent prostate cancer suggest sensitivity of 94–98% for the identification of a prostatic cancer origin, although these numbers might represent a slight underestimate because of an overrepresentation of Gleason 4 + 4, 5 + 5 and recurrent prostate cancers in our cohort. Accordingly, the sensitivity of PSAP (96.5%) and PSA (99.8%) were slightly higher in previous studies of our group analyzing large consecutive prostate cancer cohorts including much higher proportions of Gleason 3 + 3 and 3 + 4 cancer than in the current set of tumors. The specificity for the distinction of prostate cancer was somewhat lower for prostein (91.7%) as compared to the 100% for PSAP and PSA (99.7%) observed in these earlier studies [20, 21]. However, the characteristic granular perinuclear staining pattern that can hardly result from staining artefacts is a major strongpoint of prostein IHC which may thus justify the use of prostein antibodies as a part of a diagnostic panel for the identification of a prostatic cancer origin.

The location of the prostein protein in subcellular vesicles in the cytoplasm and co-localization to other compartments, i.e., the endoplasmatic reticulum fits well with the estimated function of prostein as a sucrose transport protein [2, 22]. However, many of the extra-prostatic tumor entities that were most commonly prostein positive were adenocarcinomas or neuroendocrine tumors. As these cell types share a secretory or neurosecretory function it might be speculated that prostein may have also a general role in cell secretion. The comparison of detectable prostein expression with histopathological and molecular tumor parameters in breast, colon, gastric and pancreatic adenocarcinoma had revealed only few statistically significant associations which do not provide strong evidence for a relevant biological/clinical role of prostein in non-prostatic cancers. It is possible that these findings represent statistical artifacts attributed to the high number of statistical analyses executed in this study.

Considering the large scale of our study, our assay was extensively validated by comparing our IHC findings in normal tissues with data obtained by another independent anti-prostein antibody and RNA data derived from three different publicly accessible databases [22,23,24,25]. To ensure an as broad as possible range of proteins to be tested for a possible cross-reactivity, 76 different normal tissues categories were included in this analysis. The validity of our assay was supported by the finding of the highest levels of prostein immunostaining in the prostate, the organ with the highest documented RNA expression level and the finding of prostein positive cell populations in most other organs with documented low level RNA expression such as the stomach, respiratory epithelium, hypophysis, spleen, and the brain. Only RNA expression in the liver could not be corroborated by our assay. That all prostein positive cell types detected by MSVA-460R (islet cells of the pancreas, respiratory epithelium, epithelial cells of the adenohypophysis, surface epithelial cells of the stomach, glia cells in the brain, monocytic cells in the spleen and lymph nodes) were also identified by the independent second antibody EPR4795(2) (Supplementary Fig. 1) adds further evidence for the validity of our assay. Additional stainings of the placenta and the testis which were only observed by EPR4795(2) were considered antibody specific cross-reactivities of this antibody and suggest that this antibody is less appropriate for prostein assessment.

Conclusion

Our data provide a comprehensive overview on prostein expression in human cancers. The data show that prostein is a highly sensitive prostate cancer marker with positive results in at least 98% of primary prostate cancers. Because prostein can also be expressed in various other tumor entities, the classification of a tumor mass as a prostate cancer should not be made based on prostein positivity alone.