Overexpression of Heat Shock Protein 90α in Canine Mammary Tumors

Himalaya Bhardwaj Guru Angad Dev Veterinary and Animal Sciences University B V Sunil Kumar (  drbvskumar@rediffmail.com ) GADVASU: Guru Angad Dev Veterinary and Animal Sciences University Chanchal Singh Guru Angad Dev Veterinary and Animal Sciences University Kuldip Gupta Guru Angad Dev Veterinary and Animal Sciences University Ashwani Kumar Guru Angad Dev Veterinary and Animal Sciences University Digvijay Singh Guru Angad Dev Veterinary and Animal Sciences University


Introduction
Geriatric and unspayed bitches have the most common, dangerous, and frequently diagnosed neoplasm, i.e., mammary tumor. It is also life-threatening to bitches due to distant metastasis development affecting other organs of the body Misdorp 2002). The incidence of mammary neoplasia in female dogs (bitches) is approximately three times more than that in women, followed by cats and other pet animals (Birdi et  These proteins get into the circulation, are quanti ed, and are further used to diagnose the malady ). Therefore, heat shock proteins, which play a pivotal role in moderating cellular stress, represent a group of molecular markers of cancer Pandey et al. 2018).
Heat shock proteins (Hsps) are members of the molecular chaperone family that play essential roles in folding a large number of cellular proteins apart from acting as mediators of resistance to hyperthermia in all cellular organisms (Calderwood 2010). They also play an important part in cell proliferation and drug resistance, making them proteins of special clinical interest. Elevated levels of Hsps are seen in a wide spectrum of malignant tumors, including mammary carcinoma and some members viz. Hsp27, Hsp70, and Hsp90 are thought to play roles in breast cancer (Wang et al. 2009;Liu et al. 2019).
Hsp90 is one of the most abundant proteins in mammalian cells and has four major isoforms: Hsp90A (cytosolic inducible form), Hsp90B1 constitutive form (endoplasmic reticulum/ER form or endoplasmin/Grp94), and Trap (mitochondrial form). The cytosolic form can again be sub-classi ed into Hsp90AA (Hsp90α) and Hsp90AB (Hsp90β)  isoforms. The aim of research on mammary neoplasms in female dogs should be to extend the panel of tumor markers by adding new ones that could help in diagnostics and therapy. One such marker is a subtype of HSP90, Hsp90α encoded by hsp90AA1, which is induced by stress. Hsp90AA1 is overexpressed in different types of human cancers (Wang et al. 2009). There are no data on hsp90AA1 expression in canine mammary tumors.
In the present study, all the mammary gland tumor samples were collected from unsprayed female dogs (n=20) of the 6-12 years old age group with a median of 8.5 years during surgeries performed in the Department of Veterinary Surgery and Radiology, GADVASU, Ludhiana, India. Healthy canine mammary tissues were also used from age and breed-matched dogs (n=20). After removing tumors and postoperative care, the animals were discharged from the clinics and allowed to go with their owners. No treatment of any type is given to dogs prior to surgery.
For total RNA extraction, tissue samples were collected in RNAlater TM solution and stored at -20°C.
Similarly, tissue samples were collected in 10% neutral buffered formalin for histopathological and immunohistochemical analyses. Histopathology H&E staining of tissue para n sections (5µm thick) was done using the HE stain kit (Vector laboratories).
Tissue sections were also sliced and nd mounted over clean poly-L lysine (PLL) coated glass slides for IHC. According to WHO standards, the classi cation and grading of the tumors were carried out by pathologists unknowing to subject history (Goldschmidt et al. 2011).

RNA isolation and cDNA synthesis
Extraction of RNA from mammary tissues were done using TRIzol TM reagent (Life Technologies, USA). About 1µg of the isolated RNA was used for cDNA synthesis using oligo dT primers and Bio-Rad iScript cDNA synthesis kit (Bio-Rad Laboratories, USA). After that, the synthesized cDNA was con rmed using speci c β-actin primers.

Real-Time PCR (qRT-PCR)
Relative expression of HSP90AA1 mRNA in tumorous and healthy canine mammary tissues was assessed by SYBR green-based qRT-PCR. The speci c primers targeting the HSP90AA1 gene (ACCCTGACGACATCACCAAC and GGGATCAGCTCTTCGCAGTT), along with endogenous housekeeping control genes, RPS-19 (CCTTCCTCAAAAAGTCTGGG and GTTCTCATCGTAGGGAGCAAG) ) and β-actin (CCGCGAGAAGATGACCCAGA and GTGAGGATCTTCATGAGGTAGTCGG) (Timmermans-Sprang et al. 2015) and 2x SYBR Green qPCR Mix (Real Gene, US) was used for the expression pro ling. The e ciency of these housekeeping genes had previously been tested in cancer research ). Primers of all three genes were used at 0.25µM nal concentration. Annealing and extension of all the three genes were carried out at 53°C, and 72°C and dissociation curves were generated between 65°C to 95°C to assess the speci city of the amplicons.
The percent e ciencies of the PCR ampli cation for each gene were calculated as E= (10 −1/slope -1) X 100, where the slope was calculated for the semi-log regression curve plotted between log cDNA (serially

Statistical Analysis
Statistical analyses were checked using Statistical analysis software (SAS ver. 9.3). Pearson correlation coe cient (r) was calculated between H-score and 2 −δCt (qRT-PCR score) to determine any association between them. The q-scores and H-scores were also correlated using Spearman's rank correlation.
Statistical association of tumor histotypes with tumor grades, qRT-PCR score, and H score was assessed using the Kruscal Wallis test.

Results
Histopathological analysis of CMT of the 20 tumor samples, 7 were classi ed as complex carcinoma, three each were of anaplastic carcinoma, papillary carcinoma, and early carcinoma, respectively, two of mixed tumors and one of hemangiosarcoma and mast cell tumor. About 45% (n=9) were designated as grade II (moderately differentiated), 35% of tumors (n=7) were found to be of grade I (well-differentiated), and only 20% of tumors (n=4) belonged to grade III (poorly differentiated).

Expression analysis of HSP90AA1 mRNA using Real-Time PCR
In the present study, HSP90AA1, RPS19, and β-actin genes had ampli cation e ciencies of 89.15%, 94.24%, and 95.65%, respectively, suggesting that the exponential e ciencies of the HSP90AA1 gene along with the internal controls were suitable for real-time PCR. HSP90AA1 gene was found to be 4.613 ± 0.49 times signi cantly (P ≤ 0.05) overexpressed in CMT as compared to the healthy tissue glands.

Immunohistochemical detection of HSP90α expression in canine mammary neoplasia
The IgG puri ed from the hyperimmune sera of rabbit distinctly reacted with native HSP90α isolated from the tumor tissue homogenate upon Western blotting. A distinct band corresponding to 64 kDa was spotted on the nitrocellulose membrane (Fig. 1), which con rmed that the puri ed IgG speci cally reacted with the native cellular HSP90α (Bhardwaj 2022). Upon IHC, the HSP90α immunopositive cells were stained brown. The negative controls, in which unimmunized rabbit serum was used, did not show any immunostaining with the raised HSP90α antibodies. Immunoreactivity against cellular HSP90α was judged based on the pathologist's staining index (SI). It has been reported that in the current study, a mild cytoplasmic expression of Hsp90α (SI<6) was detected in the majority of healthy canine mammary tissue (95%). In contrast, a varied Hsp90α expression was evident in most CMT histotypes under study. 75% of CMT tissues showed a strong cytoplasmic immunoreactivity (SI ≥6) for Hsp90α (Fig. 2a-2h), which correlated with the aggressiveness of CMT. Weak immunostaining (SI<6) against Hsp90α was detected in two sections of papillary carcinoma and one section in hemangiosarcoma. At the same time, a strong immuno-staining (SI ≥6) was seen in one section of mixed mammary tumors, mast cell tumors, and three sections of anaplastic carcinomas. 100% sections of complex carcinomas showed strong immunoreactivity against the protein. Moreover, a strong Hsp90α protein expression was detected in 33.33% of grade-1, 80% of grade-2, and 100% of grade-3 CMT.

Relationship between HSPD1 expression at mRNA and protein level in various types of CMT
The statistical analysis between the IHC score (H-Score) and qRT-PCR score revealed that a positive and high (P<0.005) correlation between HSP90AA1 transcript and protein expression with a Pearson Correlation coe cient of 0.6289, as evident by the Scatter plot (Fig. 3). Also, a spearman's rank correlation of 0.69702 (P<0.005) was found between H-score and qPCR score, indicating that rankings for HSP90AA1 mRNA and Hsp90α protein expression were highly similar among various CMT samples. Though, the Kruscal Wallis test revealed no signi cant differences in the H scores with respect to different CMT histotypes indicating that Hsp90AA1 was overexpressed in CMT irrespective of the histotypes.

Discussion
Canine mammary tumors (CMTs) are a dangerous condition in the modern era and are becoming more prevalent globally. Nevertheless, with the advent of advanced technology, it is possible to diagnose most of these tumors at an early stage (Pandey et al. 2018). Diagnosis of mammary tumors at an early stage is as important as in later stages, and they may metastasize. Diagnosis, if done via histopathology, involves invasive techniques of tissue collection. Milder steps like ne needle aspirate (FNA) are less invasive but less accurate also. As it is di cult to determine the biological behavior of mammary tumors by microscopic examination, markers are important tools in its diagnosis. With the advancement of research in cancer, various biomarkers have been successfully discovered for a different type of human neoplasia. Heat shock proteins are one of the potential biomarkers in human carcinogenesis ). HSP90α is a heat shock protein that also plays an indispensable role in immune responses and apoptosis.
Several reports advocate that HSP90α is upregulated in the different types of human cancer like breast cancer, liver, head, and neck cancer. Given the tight relationship between CMT and human breast cancer, we decided to investigate HSP90α expression in canine mammary cancers. Like our present study, many reports also observed that complex carcinoma, among all other canine mammary tumors, is the most persistent (Birdi et  Likewise, overexpression of the HSP90AA1 gene has been found in human breast cancer patients, related to poor prognosis and less survivability. (Klimczak et al. 2019). Yano M et al (1999) investigated the expression level of hsp90α, hsp90β, and cyclin D1 in human breast cancer. Breast cancer tissues were found to have signi cantly higher levels of mRNAs coding for hsp90α and cyclin D1 than non-cancer tissues. Previous studies have revealed over-expression of hsp90α also found in the different types of tumor-like in pancreatic carcinoma cells (Gress et al. 1994;Yano et al. 1999) and leukemic cells (Yano et al. 1999;Yufu et al. 1992). Detection of Hsp90α has been considered a poor prognostic marker of human breast cancer (Jameel et al. 1992;Yano et al. 1999). Also, one recent nding showed increased hsp90α in head and neck cancer (Fan et al. 2020). Our outcomes were consistent with these results, and we also reported overexpressed hsp90α in canine mammary tumors.
Different types of HSPs were also upregulated during the malignant alteration of mammary glands in pets. A recent study showed that Cytoplasmic Hsp90 protein level was signi cantly (p < 0.0001) higher in ductal carcinoma in situ and aggressive breast carcinomas than normal breast tissue. (Diehl et al. 2009). It has been reported that with a higher stage of mammary gland tissue malignancies, HSP90 expression has been correlated in dogs. The expression was highest in solid and complex carcinomas (Badowska-Kozakiewicz and Malicka 2012). Immunohistochemistry has also revealed that this isoform of HSP90 is overexpressed in distinct histotypes of canine mammary cancers. Grp94 was overexpressed in all forms of canine mammary cancers examined, regardless of histological types and grades , which was similar to our present study in which Hsp90α, another isoform of heat shock protein, overexpressed in CMT histotypes. We are rst to report an elevated level of Hsp90α expression in various histotypes of canine mammary tumors. Hence, adding Hsp90α to the current panel of CMT biomarkers will certainly improve the sensitivity and speci city of CMT diagnosis. However, owing to the lesser sample size, the data need to be con rmed in a study employing a large number and types of CMT in the future.

Con ict of interest:
None of the authors of this paper has a nancial or personal relationship with other people or institutions that could inappropriately in uence or bias the content of the paper.