3. KCNQ1, TP53, and IL2RG Genetically Engineered Hamster Cancer Models
3.1. KCNQ1 Knockout Hamster Model
KCNQ1 is a developmentally imprinted potassium channel gene that is widely expressed in both human and rodent tissues. Among other functions, KCNQ1 is well known for its voltage-gated interaction with its heterodimeric partner KCNE1 to regulate cardiac myocyte repolarization.
KCNQ1 mutations in humans cause a range of disease pathologies in humans including cardiac arrhythmia, inner ear defects, and gastric hyperplasia. Notably, work by the group has shown that
KCNQ1 acts as a tumor suppressor gene in the gastrointestinal tract in both humans and mice
[27][28]. For example, the maintenance of the expression of KCNQ1 was associated with a significant survival advantage in human colorectal cancer patients with stage IV metastatic disease
[28], and very recently it was reported that KCNQ1 was a major CRC prognostic predictor classifier as CRC patients with stage II and stage III CRC who maintained expression of KCNQ1 showed a much stronger disease-free survival (DFS)
[29].
Employing CRISPR/cas9 gene targeting an 11-bp insertion in the hamster
KCNQ1 gene was introduced into the hamster genome, which resulted in a constitutive null allele
[30].
KCNQ1 homozygous KO hamsters demonstrated similar neurological defects and other phenotypes that are similar to
Kcnq1 KO mice
[28], including inner ear defects, head bobbing and smaller stature
[30]. Adult
KCNQ1 homozygous KO hamsters developed severe physical stress as early as 70 days of age, including overt cancers at necropsy
[30]. Overall, >85% of the
KCNQ1 homozygous KO hamsters developed cancers, with the mean age when they became moribund at 150 days. None of the hamster littermate siblings that were either wild-type or heterozygous for
KCNQ1 mutations developed cancers. The four most common cancers observed following pathological analysis were T-cell lymphomas (the most common cancer), plasma cell tumors, hemangiosarcomas, and myeloproliferative disease consistent with myeloid leukemia
[30].
Figure 1 depicts gross and microscopic images of an intestinal T-cell lymphoma.
Figure 1. Gross and microscopic images of an intestinal T cell lymphoma in a
KCNQ1 homozygous mutant hamster. Grossly (
A), extending from the intestinal wall through the serosa and mesentery, there are multifocal to coalescing white-beige masses (arrow heads), with focal areas of necrosis and hemorrhage (arrows). The tumors circumferentially surround the intestines (indicated by asterisks) which are occasionally moderately dilated and segmentally ulcerated (transmural). Histologically, the tumor consists of densely packed sheets of round to oval cells with scant basophilic cytoplasm (
B) that infiltrate the intestinal mucosa ((
C), asterisk) and subjacent layers through to intestinal serosa. The cells show a uniform, intense immunolabeling of plasma membrane for CD3 (
C,
D) (anti-CD3 antibody, DAKO/Agilent, Santa Clara, CA, USA; Cat# A0452). H&E, 40× objective Panel (
B); CD3 immunolabeling with 20× objective Panel (
C), and with 40× objective Panel (
D) (from
[30], with permission).
The cancer phenotype in
KCNQ1 mutant hamsters was unique in several aspects. First, the
KCNQ1 KO hamsters represented the initial creation of a genetically engineered hamster cancer model. Second, while
Kcnq1 mutant mice demonstrated an enhanced intestinal cancer phenotype when introgressed into the
ApcMin model of intestinal cancer
[28],
Kcnq1 mutations alone in mice did not generate cancers in any tissue, unlike hamsters, who developed a range of cancers, sometimes multiple synchronous cancers, in different tissues. Similarly, humans who are mutant for
KCNQ1 do not develop cancer, aside from the development of gastric hyperplasia which can represent a premalignant stage of gastric cancer. Thus, for reasons that remain unknown, the complete loss of KCNQ1 function uniquely leads to widespread blood cell cancer development in hamsters. How these cancers arise in
KCNQ1 KO hamsters is an area of active investigation. Studies in humans and mice have shown that KCNQ1 is expressed in several types of hematopoietic cells, including thymic T cells, bone marrow, and other white blood cells, although KCNQ1 expression patterns in these cell types in hamsters have not been adequately characterized. Work by the group and others has indicated that KCNQ1 may be involved in the regulation of tissue stem cell transformation. This has been proposed in GI cancers associated with bidirectional Wnt/β-catenin signaling. Work by the group has shown that
Kcnq1 mRNA
[28] and protein expression (P. Scott et al., unpublished) was localized to the intestinal stem cell compartment. Further,
Kcnq1 expression has been reported in a murine bone marrow-derived stem cell progenitor population. Ongoing studies are exploring whether Wnt/β-catenin signaling is dysregulated in
KCNQ1-deficient bone marrow-derived hamster progenitor cells, of particular importance with the emerging evidence of a significant role for Wnt signaling in hematopoietic cancers.
3.2. TP53 Knockout Hamster Model
TP53 is considered the canonical tumor suppressor gene, commonly mutated in most major human cancers, and for this reason, it has been referred to as the “guardian of the genome”
[31]. TP53 is a sequence-specific transcription factor that is involved in a range of cellular functions that are critical for its role in suppressing cancer, including cell cycle regulation, sensing of DNA damage, sensing and reaction to oxidative stress, apoptosis, senescence, and cellular metabolism
[32]. Employing CRISPR/cas9 genetic engineering researchers created a 1-bp insertion in the hamster
TP53 gene at amino acid position 311 that resulted in a truncation mutation that disrupts the DNA binding domain of TP53
[33]. Researchers found that both
TP53 homozygous and
TP53 heterozygous KO hamsters developed a wide range of cancers starting at 53 days of age, with individual hamsters often manifesting multiple synchronous cancers in different tissues. In contrast, a control group of
TP53 wild-type siblings did not develop cancer, even when aged beyond one year. On average,
TP53 homozygous mutants survived 139 days, and
TP53 heterozygous mutants survived 286 days. Researchers analyzed by histopathology 52
TP53 homozygous mutant hamsters and 30
TP53 heterozygous mutant hamsters
[33]. The homozygous mutant hamsters developed a wide range of cancers, with lymphomas (29%), hemangiosarcomas (27%), and myeloid leukemias (21%) representing the most common cancers. To a lesser extent, they developed several other sarcomas and adenocarcinomas in the adrenal glands, pancreas, and kidney. In
TP53 heterozygous hamsters, lymphomas (67%) were the predominant type of cancer, followed by hemangiosarcomas (17%), myeloid leukemias (6%), and several other sarcomas and carcinomas that were found in only one hamster. Notably, in a group of 17 cancers from
TP53 heterozygous animals all showed loss of heterozygosity (LOH) at the
TP53 locus
[33]. See
Figure 2 depicts myeloid leukemia involvement in the liver of
TP53 mutant hamsters.
Figure 2. Myeloid leukemia involvement of liver of TP53 homozygous mutant hamster. There are multifocal infiltrates of neoplastic cells predominantly involving the portal triads ((
A), H&E, 4× objective lens). Infiltrating cells surrounding hepatic portal veins are strongly myeloperoxidase positive ((
B,
D), 20× and 40× objective lens, respectively). (
C) shows corresponding negative control (MPO-specific primary antibody is substituted with normal rabbit immunoglobulin on a serial step section, 40× objective lens) for image in (
D) (from
[33], with permission).
The cancer phenotypes in TP53 mutant hamsters were notable in several areas. First of all, in contrast with
Trp53 homozygous KO mice, hamsters developed carcinomas in several epithelial tissues that were not observed in
Trp53 KO mice. Human Li Fraumeni patients who carry germline
TP53 mutations typically develop carcinomas while
TP53 null mice primarily develop sarcomas.
Trp53 mice carrying gain of function point mutations do develop a broader range of cancers that is closer to human cancers with
TP53 mutations and
TP53-deficient hamsters with one major exception in that they do not develop myeloid leukemias, which are common in
TP53 mutants (especially homozygous mutant) hamsters and in human myeloid leukemia patients.
TP53-deficient acute myelogenous leukemia (AML) is a particularly aggressive sub-type of AML with a median survival of 5–9 months and one-year survival of 0–10%
[34]. Further,
TP53 mutations are common in therapy-related AML (t-AML), a very aggressive disease because of the high likelihood of treatment failure.
TP53 mutations are also found in 20–27% of AML with myelodysplasia-related changes (AML-MRC) and up to 40% of therapy-related AML myelodysplastic syndrome (AML/MDS)
[34].
TP53 mutations are also associated with high-risk MDS and rapid transformation to secondary AML (sAML), an extremely lethal disease with less than 6 months median overall survival
[34]. There are currently few treatment options for
TP53-deficient AML and one of the major barriers to the development of new therapies has been the lack of a reliable rodent model.
Trp53-deficient mice do not develop AML, but now there is a hamster model that can be used to study the genetic steps that cause this disease, including high-risk MDS that can rapidly transform to secondary AML. Hamsters can now also be used to test new drug modalities to treat
TP53-deficient AML.
3.3. IL2RG KO Hamster Model
IL2RG encodes the commonly shared gamma chain subunit in the interleukin-2 receptors (IL2Rs) which meditate the functions of multiple type I interleukin cytokines, including interleukin-2 (IL-2), IL-4, IL-7, IL-9, IL-15 and IL-21
[35][36].
IL2RG resides on the X-chromosome and loss of function mutations in this gene cause X-linked severe combined immunodeficiency (XSCID) which is characterized by a failure in T development, lack of class switching recombination of immunoglobin genes in B lymphocytes, and the absence of natural killer (NK) cells. These severe immunodeficiency features highlight the importance of IL2RG in the development of adaptative immunity and innate immunity.
To develop a hamster model for XSCID, as well as a host for human tumor engraftment (discussed below), researchers recently created several
IL2RG KO hamster lines. To genetically inactivate
IL2RG in the hamster, researchers employed the CRISPR/cas9 system with single guide RNAs (sgRNAs) designed to target its exon 1. Among the produced hamster lines carrying different insertions and deletions (indels) in exon 1, lines with frameshift mutations causing multiple premature stop codons in
IL2RG were chosen for the establishment of
IL2RG KO hamster colonies and used to produce experimental animals
[37][38]. Lymphocyte-specific gene expression analyses in the spleen showed that
IL2RG KO hamsters present severe defects in lymphocyte development which are characterized by a greatly reduced number of CD4
+ T cells and barely detectable CD8
+ T cells, B cells, and NK cells, mimicking many aspects of human XSCID immunodeficiency. The use of these
IL2RG KO hamsters as an XSCID model to study the infectious diseases that are of great health concern in immunocompromised patients was reported
[36]. Here, researchers summarize the recent studies in using the
IL2RG KO hamsters as a host for human tumor tissue engraftment to produce patient-derived xenograft (PDX) models as laboratory avatars for oncology research.
The use of immunodeficient animals, either with IL2RG-deficiency alone or in combination with other immunodeficiencies such as RAG2-deficiency, as hosts to produce PDX models relies on their incapability of mounting immune rejections to human tissues, therefore allowing them to be xenografted in the animal hosts. Compared to in vitro cultured tumor-derived cell lines where in vitro adaptation takes place which may change the physiology of tumors, the PDX models better recapitulate the tumor biology in vivo
[38].
Il2rg KO mice have been widely used as hosts for producing PDX models
[39]. However, there are several significant limitations in using mice as hosts for human tissue xenografting, chief among which are the incompatibility of mouse growth factors and cytokines with human tissues and cells, leading to low xenografting rates, limited or no self-renewal of malignant stem cells, and low cancer cell metastasis rates. The lack of functional communications between human cells/tissues and the mouse physiological milieu is also one of the underlying causes of the fundamental problems in using immunodeficient mice as hosts for human hematopoietic stem cell transplantation to produce so-called humanized mice carrying a humanized hematolymphoid system. Due to the lack of functional support from mouse cytokines for human lymphoid and myeloid cells, mice are incapable of fully supporting the development of a human immune system, as manifested by impaired lymph node development, poorly developed germinal centers, defective humoral immune responses (characterized by low levels of immunoglobulin production and inefficient immunoglobulin class switching recombination) and human NK cell development
[40][41][42][43][44]. To address this species incompatibility issue, next-generation humanized mice transgenically expressing human cytokine genes, such as
IL-3 and granulocyte macrophage-colony stimulating factor (
GM-CSF) that are critical for myeloid cell development
[45], and other human genes such as in the MISTRG mice
[46], have been developed. The transgenic expression of these human genes has significantly improved the engraftment of human tumor tissues for mouse PDX model production
[47] and the development of human lymphoid and myeloid cells. Nevertheless, these humanized mice are still not optimal in several aspects including being defective in mounting antigen-specific adaptive immune responses, low levels of reconstitution of gut-associated lymphoid tissues from human cells, and poor lymphoid architecture and organ development
[42].
Therefore, a small rodent with better physiological compatibility with human cells/tissues is highly desired as a host for PDX and immune system humanization. In this regard, the golden Syrian hamster is a promising candidate, as in contrast to mice several human cytokines such as GM-CSF
[48] and IL-12
[49][50] are cross-reactive with hamster cells. As mentioned above, human GM-CSF is among the key cytokines transgenically expressed in the MISTRG mice, which greatly improved human myeloid cell reconstitution
[45]. To test whether the hamster is indeed a suitable host for human PDX, researchers carried out human pancreatic cancer cell transplantation experiments in the
IL2RG KO hamsters and produced the first hamster PDX model. In these experiments, researchers used an
IL2RG KO line (named ZZU001) that researchers generated in which a 10 bp frameshift deletion in exon 1 fully inactivated the expression of the
IL2RG gene
[37].
Researchers first performed subcutaneous xenotransplantation of a human pancreatic cancer cell line, MIA-PaCa-2
[51], into ZZU001 hamsters, and for comparisons, also into immunodeficient B-NDG (NOD-
Prkdcscid IL2rgtm1/Bcgen) mice which are defective in both the
IL2rg gene and the
Prkdc gene. While xenografted human pancreatic cancer cells developed in both the hamster and B-NDG mouse hosts, no distant metastasis was identified in mice but was found in the lung, kidneys, and adrenal glands in ZZU001 hamsters.
Figure 3 (From
[37]) shows a H&E staining of a lung metastasis at 22 d, 29 d, and 36 d after subcutaneous injection of MIA-PaCa-2 cells into ZZU001 hamsters and B-NDG mice. The same observations, i.e., metastasis was only observed in the hamster hosts but not in the mouse hosts, were also made from each of the other four independent human pancreatic cancer cell lines tested.
Figure 3. (From
[37], with permission). H&E staining of a lung metastasis at 22 d, 29 d, 36 d after subcutaneous injection of MIA-PaCa-2 cells in ZZU001 hamsters and B-NDG mice. Scale bars = 100 µm.
Because orthotopic transplantation models mimic the biology of primary tumors more closely than subcutaneous transplantation models, to establish the feasibility of the hamster as an orthotopic model for human pancreatic cancer cell transplantation, orthotopic transplantation experiments with MIA-PaCa-2 cells were also performed in both ZZU001 hamsters and B-NDG mice. MIA-PaCa-2 cells were directly injected into the pancreas of hamsters and mice and tumor formation rates, as well as tumor metastasis, were compared between these two host species. Orthotopic transplantation of Mia-PaCa2 cells in hamsters led to a 100% (5/5) tumor formation rate and metastasis at multiple sites, including liver (100%), lung (100%), retroperitoneum (100%), mesentery (100%), kidney (100%), diaphragm (40%), adrenal gland (40%), and stomach (20%). In contrast, much lower metastatic rates to other organs were observed, with no lung metastasis observed at all, in the B-NDG hosts. In addition, metastatic tumors developed in the hamster orthotopic models presented similar clinical and pathological features to what was observed in human patients, which includes local infiltration of cancer cells, ascites, jaundice, ileus, and cachexia.
In summary, researchers demonstrated that in comparison to the highly immunodeficient B-NDG mice, ZZU001 hamsters with the loss of IL2RG function alone (therefore much less immunodeficient than B-NDG mice) support better human pancreatic cancer cell engraftment and multiple organ metastasis and serve as an improved host for human pancreatic cancer engraftment. Researchers posit that the improved engraftments and higher metastasis rates in the ZZU001 hamsters than in mouse hosts by human cancer cells are attributed to better communications between the human cancer cells with the hamster physiological milieu.