In vivo Expression Analysis of Xenografts in Severe

Xenotransplantation was performed with infected cells derived from both breast and non-breast cells. To determine whether or not thein vitroß-galactosidase/EGFP expression of the MMTV-BAGgal hybrid vector is mirrored in anin vivosetting, human tumour cells that had been stably infected with the hybrid vector were taken up in matrigel before being injected into SCID mice.

The use of matrigel embedded cells has been described as being particularly useful when transplanting human breast tumours cells. It has been observed that especially breast tumour cells injected in the presence of matrigel exhibited close histopathalogical resemblances to surgical breast specimens after the tumours that arose had been recovered. Co-injection with matrigel can increase the incidence of tumours and enhanced the tumour growth. Estrogen supplementation was found also to increase the growth rate of such tumours (Rae-Venteret al., 1980; Osborneet al., 1985). An estrogen pellet (1.7 mg) was therefore transplanted into the

SCID/bg mice 24 hours before the transplantation took place (see 2.6.1). Such pellets release estrogen for a period of approximately 90 days.

3.1.11.1 Transplantation of Infected Breast and Non-Breast Cells into SCID/bg Mice

Human breast or non-breast tumours cells that had previously been infected with either the MMTV-BAGgal hybrid vector or BAGgal were cultivated in a monolayer cell culture system.

1x10⁷cells per SCID mouse were prepared for the implantation (see 2.6.1.1). These cells were mixed together with matrigel on ice and immediately injected into the mammary fad pad (mfp) of SCID/bg mice. 4 mice were injected per construct and cell line. The mice were then examined 2 times per week and any tumours measured. 5 to 8 weeks later tumours could be observed in all treated mice, the time depending on the cell line used. The mice were sacrificed when the tumour had reached a size of 10 to 15 mm. The tumour was removed from the surrounding tissue and both RNA and protein prepared for further analysis (Fig. 3.26).

Figure 3.26: Transplantation of Infected Breast and non Breast Cells into SCID/bg mice Mice were housed in microisolator cages, fed only autoclaved food and water, and handled only with gloves.

Female mice that weighed between 17 and 20 g were anaesthetised with ketamin (Ketalar 1%, 0.2 ml/20 g mouse i.p). Antibiotic (Borgal^R24%) was added to the drinking water on the day of implantation. For the implantation, a small area of skin was shaved (~2x2 cm) and disinfected. A small incision was made in the skin and the oestrogen pellet implanted. Depending on the size of the wound, the cut was either sewn together (by a veterinarian) or simply closed. 24 hours later 1x10⁷ cells that had been previously mixed with Matrigel were injected into the mammary fat pad. After a 5-8 week incubation period, the mice were killed and the tumour immediately removed from the surrounding tissue. A SCID/bg mouse with tumour derived from primary mammary tumour cells can be seen.

3.1.11.2 Protein Expression Analysis

Protein was extracted from the tumours of 3 mice per tumour type. Tumours were homogenised on ice in lysis buffer (see 2.5.3.1). The tumour extracts were then treated with a cocktail of protease inhibitors [PMSF (0.2 mM) and leupeptin (5 µg/ml)] in conjunction with a heat inactivation step, just before analysis, in order to reduce high levels of endogenous ß-galactosidase activity that had been previously observed in tissue samples (Shaperet al.,1994).

Every human breast tumour showed high levels of ß-galactosidase expression from the MMTV-BAGgal hybrid vector as compared to the parental MMTV-BAGgal (Fig. 3.27a).

In vivobreast tumours

MCF7 ZR-75-1 T-47D

0 10 20 30 40 50 60 70

Mock BAGgal

MMTV-BAGgal 150200

250 300

foldbackground

In vivohuman non-breast tumours

EJ Panc-1

0 1 2 3 4 5 6 7

Mock BAGgal

MMTV-BAGgal

foldbackground

a)

b)

The T-47D cells showed expression levels from the MMTV-BAGgal hybrid vector approximately 250 times that of the background, whereas the BAGgal levels were, on average only 1.4 times that of the empty cells. This high level of expression could also be shown histologically in both T-47D and ZR-75-1 tumours (Fig. 3.28a) after fixation (see 2.5.3.2.2) and staining.

Figure3.27: Analysis of ß-galactosidase Expression in Breast and non-Breast Tumour Cell Xenografts in SCID/bg mice

a) Expression of ß-galactosidase in human breast cell derived tumours in SCID/bg mice.

10 µg of protein that had been previously been extracted from tumours that had arisen after implantation of tumour cells was used for a quantitative galactosidase assay (Galactolight, Perkin Elmer). The expression of galactosidase is measured in relative light units after a chemoluminescent substrate is added. The background ß-galactosidase activity was set to a value of 1 and the expression of the vectors then calculated in relationship to the background. High levels of expression could be seen from the hybrid MMTV-BAGgal construct in MCF7, ZR-75-1 and T-47D cells. Expression from the BAGgal vector was much lower by comparison.

b) Expression of ß-galactosidase in human non-breast cell derived tumours in SCID/bg mice

The experiment was performed in the same manner as a). Expression of both the hybrid MMTV-BAGgal vector as well as the parental BAGgal could be seen although the levels of expression are much lower than observed in the breast cell derived tumours. No significant expression could be seen in Panc-1 cells. The absolute levels of expression from both BAGgal and MMTV-BAGgal vectors were much lower in both EJ and Panc-1 cells derived tumours than any of the human breast cell derived tumours, with the exception of BAGgal vector in T-47D cells.

2 1

2 1

T-47D Tumour

non-infected MMTV-BAGgal

Panc-1 Tumour

a)

b)

In the non-breast tumours cell lines examined, EJ and Panc-1, the expression levels from the MMTV-BAGgal hybrid vector were much lower than those observed in the breast cell derived tumours (Fig. 3.27b), the highest levels being obtained in EJ cells (~6x background for MMTV-BAGgal). No significant expression could be seen in Panc-1 cells (Fig. 3.28b) Histological analysis showed similar results, with only very low expression detectable in EJ cells (not shown).

Figure3.28: Histological Analysis of ß-galactosidase Expression in Breast and non-Breast Tumour Cell Xenografts in SCID/bg mice

a) X-gal staining of breast cell derived tumours.

A T-47D derived tumour was dissected and cut into small pieces before being fixed. In 4% paraformaldehyde for 24 hours at 4°C. The tissue was then stained with x-gal (2.5.3.2.2) before being embedded in paraffin and photographed. Blue staining indicates areas of ß-galactosidase expression. Similar results were also obtained for ZR-75-1 tumours.

b) X-gal staining of non-breast cell derived tumours.

The experiment was performed in the same manner as a) No expression, normally indicated by blue areas, can be seen in the Panc-1 derived tumour.

3.1.11.3 RNA Expression Analysis

RNAanalysis was performed to show the expression levels of ß-galactosidase in the tumours arising from breast and non-breast derived cells. Northern blot analysis was made as it not only indicates the levels of expression but also allows the length of the transcript to be shown. Total RNA was isolated from a tumour from 1 of the 4 mice per construct per cell and run over a denaturing formaldehyde gel (see 2.2.2.3.2) and, after being transferred to a nylon membrane via capillary blotting (see 2.2.7.2), was hybridised to a radioactively labelled 0.4 kb ß-galactosidase specific fragment. The expected 7 kb fragment that could be seen in all human breast cell derived tumours infected with both the hybrid MMTV-BAGgal as well as the parental BAGgal vector (lanes 1, 2, 3, 6, 7 and 8) corresponding to the full length vector transcript. The BAGgal vector does, in fact, give a slightly shorter transcript than the MMTV-BAGgal vector. This is due to the presence of the larger MMTV U3 region in the 3' LTR, which makes the transcript approximately 700 bp longer (Fig. 3.29). The strongest ß-galactosidase expression could be observed in those human breast cell derived tumours that had been infected with the hybrid MMTV-BAGgal vector before implantation. The ß-galactosidase expression could also be detected in bladder carcinoma cell derived tumours (EJ, lanes 4 and 8) where the MLV promoter showed higher levels of expression than the promoter of MMTV. However, in pancreatic cell derived tumours (Panc-1, lanes 5 and 9), no expression could be detected from either the MMTV-BAGgal or BAGgal vectors. The same membrane was then stripped and hybridised to a radioactively labelled 1.2 kb ß-actin specific fragment. It shows the presence of similar quantities of a 2.2 kb ß-actin specific transcript in each lane.

Figure 3.29: Northern Blot Analysis of Human Tumours from SCID/bg Mice

20µg of total RNA was separated on a denaturing formaldehyde gel transferred to a nylon membrane using capillary blotting. The membranes was then hybridised to anα³²P labelled 0.4 kb ß-galactosidase specific fragment (from pBAGgal afterMluI digestion) and, following documentation and stripping, also to anα³²P labelled 1.2 kb ß-actin fragment [from pAL41 (Alonsoet al., 1986) afterPstI digestion]. Lane 1, RNA from an MMTV-BAGgal infected ZR-75-1 tumour; lane 2, RNA from an MMTV-BAGgal infected T-47D tumour; lane 3, RNA from an MMTV-BAGgal infected MCF7 tumour; lane 4, RNA from an MMTV-BAGgal infected EJ tumour; lane 5, RNA from an MMTV-BAGgal infected Panc-1 tumour; Lane 6, RNA from a BAGgal infected ZR-75-1 tumour; lane 7, RNA from a BAGgal infected T-47D tumour; lane 8, RNA from a BAGgal infected MCF7 tumour; lane 9, RNA from a BAGgal infected EJ tumour; lane 10, RNA from an MMTV-BAGgal infected Panc-1 tumour. Lanes 11-15 represent non-infected cells.