Characterising Mutational Spectra of Carcinogens in the Tumour Suppressor Gene TP53 Using Human TP53 Knock-in (Hupki) Mouse Embryo Fibroblasts

DNA in dividing cells is prone to mutagenesis, with mutations making key contributions to human disease including cancer. The tumour suppressor gene TP53 is the most frequently mutated gene in human tumours. Here, we present a robust protocol for studying TP53 mutagenesis utilising human TP53 knock-in (Hupki) mouse embryonic fibroblasts (HUFs). In the HUF immortalisation assay (HIMA), primary HUFs are treated with known or suspected carcinogens at 3% oxygen and then transferred to 20% atmospheric oxygen to induce senescence. Cells containing mutations (e.g., in TP53) that allow bypassing of senescence eventually emerge as immortalised clonal cell lines after 2–3 months of serial passaging. As not all immortalised HUF cells contain TP53 mutations, we developed a Nutlin-3a counter-screen to select for TP53-mutated clones prior to sequencing. TP53 mutation spectra generated can be compared with those of human tumours recorded in the International Agency for Research on Cancer TP53 mutation database. Environmental mutagens that have demonstrated and validated the utility of the HIMA include ultraviolet radiation, aristolochic acid, and benzo[a]pyrene. The TP53 mutation patterns induced by these mutagens in the HIMA corresponded to those found in human tumours from patients exposed to these mutagens. The approach presented helps to deepen our understanding of human cancer aetiology.


Introduction
The transcription factor p53 is usually kept at low levels in normal, unstressed cells, but it is stabilised and activated in response to certain stresses (e.g., DNA damage) [1]. This activation leads to a variety of outcomes such as cell cycle arrest, senescence or apoptosis depending on the severity of the damage. By preventing the growth of stressed and damaged cells, p53 plays a vital role in tumour suppression [2]. The function of p53 can be altered by mutations in the TP53 gene that encodes for p53. TP53 is the most commonly mutated gene in cancer with around 50% of all human tumours harbouring a mutation in TP53. These are mostly missense mutations occurring in the DNA binding domain encoded by exons 5-8 [3]. The International Agency for Research on Cancer (IARC) curates a database (www.p53.iarc.fr) of all mutations found in the TP53 gene published in the scientific literature. This database currently lists around 30,000 mutations in human cancers. Most missense mutations in TP53 cause a loss of function such that tumour suppressor capability is lost. However, some TP53 mutations can lead to a gain of function, whereby the mutant p53 acquires a new activity [4].
A unique tool to study carcinogen-induced human TP53 mutations in a mammalian cell context uses Hupki mouse embryo fibroblasts (HUFs) to perform the HUF immortalisation assay (HIMA). The Hupki mouse contains a partial human TP53 knock-in allele, in which exons 4-9 of the murine Trp53 gene have been replaced by the corresponding human exons, where most TP53 mutations are found in human tumours ( Figure 1) [5]. The p53 protein of the Hupki mouse functions normally and the mice are not cancer prone, unlike Trp53 knockout mice which develop tumours (mostly lymphomas) at 3-6 months of age [5,6]. The key advantage of mouse embryo fibroblasts (MEFs) is that they undergo p53-dependent senescence after around 5-6 population doublings under normal culture conditions (37 °C, 20% O2, 5% CO2) [7,8]. MEFs can bypass senescence by a disruption of either the retinoblastoma or p53-protein pathway and thus, a mutation in Trp53 is sufficient to immortalise MEFs. The immortalisation of human cells requires the disruption of both pathways in addition to a halt of telomere attrition [9]. It also takes much longer as human cells only enter senescence after 50-60 population doublings under standard culture conditions. The original protocol for the HIMA was published by Liu et al. [10] (Figure 2). The assay is initiated by treating primary HUFs with a mutagen, followed by serial passaging of treated cells and untreated controls. Cultures will undergo growth arrest due to the sensitivity of MEFs to 20% oxygen. However, most mutagen-treated cultures will harbour mutated cells that are able to bypass senescence, start proliferating again and eventually become immortalised cell lines. Additionally, untreated cells can undergo spontaneous immortalisation due to mutations acquired through culture conditions (e.g., due to oxidative stress). DNA from immortalised cells can then be isolated and sequenced to identify TP53 mutations [10] (Figure 2). Up to 30% of carcinogen-treated and 0-10% of untreated immortalised cultures harbour mutations in TP53 [11][12][13][14][15], while the remaining immortalised cultures most likely have mutations in other genes related to senescence bypass [16]. The HIMA is a unique in vitro mutation assay as it assesses the mutagenesis of a human gene that plays an important role in cancer. Other in vitro mutation assays use either non-mammalian genes (e.g., lacI, lacZ, cII) or genes with no known role in cancer (e.g., HPRT, TK) [17][18][19][20].  [10]. Primary HUFs are isolated and grown under standard culture conditions (20% oxygen). For the HIMA primary HUFs are seeded on six-well plates (P1), treated with a mutagen and serially passaged under standard culture conditions (20% oxygen) until cells undergo senescence (P5+). Mutated cells eventually emerge from senescence and can be expanded into immortalised clonal cell lines (P8+). TP53 sequencing with isolated DNA from all clones is performed to identify TP53 mutations and to evaluate the pattern of mutations induced by the mutagen.

Experimental Design
Prior to initiating the HIMA, mutagen treatment conditions must be optimised to ensure that sufficient DNA damage is induced while maintaining a population of viable cells. Therefore, the cytotoxicity of the known or suspected mutagen to be tested should first be assessed to identify a desirable concentration and an appropriate treatment time. It is important to note that the assessment of cytotoxicity helps to optimise treatment conditions for the HIMA, however, enhanced cytotoxicity is not necessarily a predictor of DNA damage and subsequent mutagenicity [21]. Thus, additional screening assays can help to further guide the HIMA treatment conditions. When possible, DNA damage (e.g., pre-mutagenic DNA adducts) can be measured directly in mutagen treated cells (see Section 3.4.3). Alternatively, induction of the DNA damage response (DDR), after treatment with a variety of sub-cytotoxic and cytotoxic concentrations of a mutagen, can be assessed by western blotting or immunofluorescence for DDR proteins such as p53, p21, pChk1/2, and pH2ax (see Section 3.4.2). Additionally, we have created and isolated HUFs from the Hupki mouse with an integrated lacZ gene. The mutagenic activity of a known or suspected genotoxicant can be determined in lacZ and used as a reference point for identifying the level of mutations induced by the test agent in lacZ under different treatment conditions before starting the more laborious HIMA (see Section 3.4.4).
Over the past few years, we have developed and validated modifications to the original HIMA protocol to improve the sensitivity and applicability of the assay [13,15,22], and the updated experimental approach is illustrated in Figure 3. The two major changes are explained below.
First, we recommend culturing primary HUFs at 3% O2 instead of 20% O2 at the start of the assay (i.e., the first two weeks in culture) as well as when optimising treatment conditions. HUFs are highly sensitive to the oxidative stress that occurs when cultured under standard conditions (20% O2). Growth progressively slows until senescence occurs after approximately 2 weeks. HUFs grown at 3% O2 do not senesce and will undergo more population doublings than those grown at 20% O2, therefore culturing under lowered O2 produces more cells. Additionally, this may reduce the number of spontaneous, background mutations caused by oxidative damage to DNA. Following mutagen treatment, HUFs must then be transferred to 20% O2 to induce senescence and select for TP53 mutations. We have shown that HUFs initially cultured at 3% oxygen that were transferred after 11 days to 20% oxygen were still able to senesce and become immortalised in the same timeframe as HUFs continuously cultured at 20% oxygen. Additionally, using this method, the percentage of spontaneous TP53 mutations has been proven to be very low (0-4%). oxygen. Prior to the HIMA treatment conditions must be optimised. For this purpose, cells are treated with the mutagen of interest and cytotoxicity, DNA damage response, genotoxicity (i.e., DNA adduct formation) and/or lacZ mutagenicity are used to determine treatment conditions for the HIMA. For the HIMA primary HUFs are seeded on six-well plates (P1) and treated with the test agent using the optimised treatment conditions. Treatment is conducted at 3% oxygen. Cells are then serially passaged under standard culture conditions (20% oxygen) until they undergo senescence (P5+). Cells will eventually emerge from senescence and can be developed into immortalised clonal cell lines (P8+). The Nutlin-3a counter-screen is used to screen immortal clones for the presence of TP53 mutations. Only TP53-mutated clones (i.e., Nutlin-3a resistant clones) will be expanded and subsequently subjected to TP53 sequence analysis. Second, we recommend performing a counter-screen of immortalised HUFs with the MDM2 inhibitor Nutlin-3a. Nutlin-3a blocks the interaction between p53 and its negative regulator MDM2 by binding to the latter [23]. When cells with WT-TP53 are treated with Nutlin-3a, p53 is stabilised and activated, leading to growth arrest [22]. Cells harbouring mutations in TP53, on the other hand, are resistant to Nutlin-3a and continue to grow normally after treatment. When following the original HIMA protocol (see Figure 2), all cultures are maintained from the first emergence of immortalised cells through the establishment of clonal cell lines, and all clones are sequenced for TP53 mutations. The majority of these clones will not harbour a mutation in TP53 but will have been immortalised by mutations in other genes. Therefore, because there is no selection specifically for TP53 mutants in the original HIMA protocol, a lot of time and effort is spent culturing clones that have WT-TP53. Our group has demonstrated that a Nutlin-3a counter-screen, as shown in Figure 3, can be used to select for mutant-TP53 HUFs and against WT-TP53 HUFs once clones emerge from senescence [22]. Most mutant-TP53 HUFs, including heterozygous TP53 mutants that retain one WT allele, are identified by the Nutlin-3a counter-screen within 2.5 months of initiating the HIMA. Only TP53-mutant cultures are expanded into cell lines and TP53-WT cells can be discarded, improving the efficiency of the assay. This allows more cultures to be included in each experiment, more assays to be performed and thus more TP53 mutants are processed [24].
The following Hupki mice are available to perform the described HIMA protocol: • The original Hupki mouse (Trp53 tm1/Holl ) on a 129/Sv background encodes arginine at the exon 4 codon 72 polymorphic site in TP53 [5]. More information can be found at The Jackson Laboratory website (www.jax.org) where this strain is available (stock no. 004301). The related strain Trp53 tm2/Holl (stock no. 008045) contains the proline variant at codon 72 in TP53 [25] • Based on the original Hupki mouse [5] we created a Hupki strain that is heterozygous for an Xpa-knockout allele (Xpa +/− ) and also harbours the pUR288 plasmid (containing the bacterial lacZ reporter gene): Hupki/Xpa/lacZ [15]. 1. For mating place one male with 1-2 females (7-8-week virgins) and check the females for a copulation plug each morning. The copulation plug usually dissolves within 12-14 h. 2. Transfer the plugged females to a new cage (=E0.5) and weigh the female on day 7, 9, 11, and 13.
CRITICAL STEP The optimal age of mouse embryos for this protocol is 13.5 days. 3. Prior to dissection (E13.5) place medium at room temperature. Place 8-10 cryogenic vials containing 1 mL 0.05% trypsin-EDTA on ice. Prepare 8-10 10-cm petri dishes with approximately 20 mL sterile PBS. Put dissection tools in a beaker containing sterile PBS. CRITICAL STEP The isolation of MEFs must occur in a sterile hood with sterile material. 4. Sacrifice the female by cervical dislocation. 5. Spray the mouse on both sides with 70% ethanol and pin down. 6. With fine tweezers or scissors, carefully cut open the mouse and pin down the skin on both sides.
CRITICAL STEP Do not touch fur with dissection tools in order to avoid contamination. 7. Cut out the string of embryos/uterus and place in 10-cm dish containing PBS. 8. Pop out each embryo from the uterus and place each in a separate 10-cm dish containing PBS. 9. Label 1.5 mL centrifuge tubes and cryogenic vials containing trypsin (from step 3) according to number of embryos and place back on ice. 10. Under the microscope remove the placenta and umbilical cord if still present. 11. Hold the embryo with sharp tweezers and remove the hematopoietic tissue with blunt tweezers. 12. Cut off the head with blunt scissors and place in a labelled centrifuge tube from step 9 and use for genotyping if necessary. 13. Place the rest of the embryo in a labelled trypsin-containing cryogenic vial from step 9 and keep on ice. 14. In between each embryo, wash dissection tools in a beaker containing PBS, then with 70% ethanol and dab on tissue to dry.

Preparation of Cells
NOTE The day of HUF isolation is defined as Day 0 (or D0) of the cells being in culture. 15. Prepare HUFs from up to four embryos at a time.
CRITICAL STEP From here onwards the preparation of HUFs must occur in a sterile cell culture biosafety cabinet. 16. Between dissociation of each embryo, rinse mincing scissors in a 50 mL centrifuge tube filled with PBS, then soak with 70% ethanol and dab on tissue to dry. 17. Use mincing scissors to mince the embryo in each vial. 18. Put up to four vials in a rack and place in the incubator at 37 °C for 10 min. 19. Pipette the embryo/trypsin mixture up and down 10× with 1 mL pipette. 20. Place the rack back in the incubator at 37 °C for a further 10 min. 21. Pipette the tissue up and down 10× with a 1 mL pipette. 22. Transfer the contents of each vial to separate 15 mL centrifuge tubes containing 9 mL of growth medium and invert. Centrifuge at 320× g for 5 min. 23. Aspirate supernatant and re-suspend the pellet in 1 mL medium by pipetting up and down 10×. 24. Add 4 mL of medium (total of 5 mL), mix, then let the tube sit for 3 min to sediment any large pieces of tissue. 25. Transfer 5 mL (avoiding pellet) from each tube to a separate 175-cm 2 flask containing 30 mL medium. 26. Move the flask back and forth and place in incubator.
CRITICAL STEP The incubator should be fitted with an oxygen sensor and a nitrogen source and set to 37 °C, 95% humidity, 5% CO2, and 3% O2. 27. Check cell growth and for any signs of contamination over the next 2-3 days. 28. Change the medium after 24 h.

Preparation of Frozen Cell Stocks
29. Label one 15 mL tube per T175 flask with the appropriate embryo number.
CRITICAL STEP It is possible to prepare frozen stocks from up to four flasks at a time. Because the freezing medium contains a high percentage of DMSO, it is important that cells are placed at -80 °C immediately to avoid any cellular damage. 30. For each embryo/flask label 5-6 cryogenic vials with ID, day in culture, passage number, portion of flask per stock and date (e.g., XE7.4, D3, P0, ⅕ T175, 01/01/2019). NOTE As described in Section 3.1.2 the day of isolation is defined as Day 0. In our experience, when following this procedure, most cultures will be 80-100% confluent three days following isolation, thus, the day when frozen stocks are prepared will be Day 3 (D3). Also, we define the preparation of frozen stocks here as Passage 0 (P0). 31. Aspirate the medium from the flask and wash with PBS. 32. Incubate with trypsin for 3 min, then add medium as indicated in Table 1, transfer to 15 mL centrifuge tube and centrifuge at 320× g for 5 min at room temperature. 33. Prepare freezing medium. 34. Aspirate the supernatant and re-suspend the pellet in 5 mL freezing medium by gently pipetting up and down. 35. Add 1 mL into each of the prepared cryogenic vials. 36. Place in a Styrofoam box at -80 °C and transfer to a liquid nitrogen tank the next day.
PAUSE STEP Frozen stocks can be stored in liquid nitrogen for many years.

Thawing of Frozen Cells. Time for Completion: 30 min Hands-on Time, 3-4 Days for Cells to Grow (D3-6/7)
37. Label two T175 flasks with embryo number, day in culture, passage number and date (i.e., XE7.4, D3, P0) and add 35 mL of medium to each flask. 38. Prepare one 15 mL tube with 9 mL medium. 39. Wearing the appropriate protective clothes, take out a vial of frozen primary cells from the liquid nitrogen tank and thaw quickly in a 37 °C water bath. CRITICAL STEP As cells were frozen in 10% DMSO it is crucial to work fast at this step to keep cells viable. 40. In the cell culture hood, pipette the thawed stock into the 15 mL tube from step 38 and centrifuge at 280× g for 5 min at room temperature. 41. Aspirate the supernatant, resuspend the pellet in 1 mL of medium and add 500 µL of cell suspension per flask. 42. To ensure an even distribution of the cells, move the flasks back and forth before placing them in an incubator set to 37 °C, 95% humidity, 5% CO2, and 3% O2. 43. The next day (D4) aspirate the medium, add 35 mL fresh medium to each flask and place flasks back into the incubator. 44. Let the cells grow for an additional 48-72 h in order to reach 80-90% confluency.

Passaging of HUFs-Time for Completion: 30-45 min for Primary Cells and up to 4 h during the HIMA
45. Once cells have reached confluency (D6-7) label the required number of cell culture dishes with culture ID, date, day and passage number and add appropriate volume of growth medium as indicated in Table 1. 46. Take flasks out of the incubator and aspirate the medium. Wash once with 12 mL PBS per flask. Add 3 mL of trypsin. Make sure the trypsin has been distributed equally and place flasks at 37 °C for 2.5 min or until cells detach. CRITICAL STEP Timing depends on the cell density and the stage of the culture (e.g., senescent cells take longer to dissociate than post-senescent cells or primary HUFs). Do not incubate cells longer than 10 min. 47. Check under the microscope if cells have detached. If cells are still attached place back at 37 °C.
Once cells have lifted, add growth medium to inactivate trypsin and transfer the cell suspension to a 50 mL centrifuge tube. 48. Count cells using a hemocytometer and seed cells at 16,000 cells/cm 2 . 49. During the HIMA, after cells have been treated with a mutagen, the cells do not have to be counted. The splitting ratio depends on how quickly the culture populates the well. Cultures should be sub-cultured at 1:1.5-1:50 splitting ratio as indicated in Table 2. For example, for a 1:3 splitting ratio pipette 500 µL of the 1.5-mL cell suspension into the correct well of the prepared 6-well plate. 50. To ensure an even distribution of the cells, move the flasks/plates back and forth before placing them in the incubator at 37 °C, 95% humidity, 5% CO2, and 3% (D0-D11) or 20% O2 (D11+). Prepare required amount of 96-well plates by labelling them with the date and passage number and by adding 100 µL PBS/well into the outside wells of the plate using a multichannel pipette. 52. Sub-culture the cells as described in Section 3.3 and place them in the incubator at 37 °C, 95% humidity, 5% CO2 and 3% O2. Prepare enough cells for 70 wells. 53. The next day prepare sufficient amounts of the highest concentration of the test compound by adding the required volume of carcinogen stock to medium. Also make medium containing an equal concentration of the solvent for the controls. 54. Make a serial dilution from the highest concentration in a 12-well reservoir. Include 6 replicates per treatment concentration. An example of treatment concentrations is shown in Figure 4. 55. Include two columns (=12 replicates) of controls. 56. Aspirate the medium from the cells using a multichannel aspirator adaptor or a multichannel pipette and dispose of accordingly. 57. Add 100 µL/well of treatment medium using a multichannel pipette and incubate for appropriate time (see Section 4.1 for advice).

Staining of Cells with Crystal Violet-Time for Completion: 30 min Followed by Drying
Step (at least 1 h) and then Another 15 min the Next Day 58. Remove the treatment medium and PBS from plates using a multichannel pipette. 59. Wash cells with 150 µL PBS/well and subsequently remove PBS with multichannel pipette. 60. Add 35 µL of 0.1% crystal violet in 10% ethanol to each well, as well as one row without cells (=blank). 61. Incubate at room temperature for at least 10 min. 62. Remove the crystal violet solution using a multichannel pipette and dispose of accordingly. 63. Wash 2× with 150 µL PBS. 64. Let the plate dry for at least 1 h shielded from light. 65. When dry add 100 µL of 50% ethanol to each well, tap to mix, and measure absorbance at 595 nm using a plate reader ( Figure 5).  134. Thaw appropriate number of vials of primary HUFs as described in Section 3.2 3-4 days prior to setting up an experiment for the DNA adduct analysis. 135. Once cells have reached confluency, label the appropriate number of 75-cm 2 flasks. 136. Trypsinise and count the cells as described in Section 3.3 and seed 1.2 × 10 6 cells per flask. 137. Place the flasks in the incubator at 37 °C, 95% humidity, 5% CO2, and 3% O2. 138. The next day treat the cells with test compound and place flask back in the incubator. 139. Once the treatment is finished, prepare a pellet as described in Section 3.6. 140. DNA adducts can be analysed by a variety of different methods [26]. In our laboratory we routinely use the 32 P-postlabelling assay to evaluate DNA adduct formation. We have published detailed step-by-step protocols previously [27,28].

Assessment of lacZ Mutagenicity (Optional)
141. Thaw appropriate number of vials of primary HUFs as described in Section 3.2 3-4 days prior to setting up an experiment to assess the lacZ mutagenicity. 142. Once cells have reached confluency, label the appropriate number of 75-cm 2 flasks (five per treatment condition). 143. Trypsinise and count the cells as described in Section 3.3 and seed 1.2 × 10 6 cells per flask. 144. Place the flasks in the incubator at 37 °C, 95% humidity, 5% CO2 and 3% O2. 145. The next day treat the cells with test compound and place flasks back in the incubator. 146. Once the treatment is finished, label the required number of 175-cm 2 flasks. 147. Trypsinise and count the cells as described in Section 3.3 and seed 2 × 10 6 cells from each 75-cm 2 flask into a 175-cm 2 flask. 148. After three days, prepare a pellet as described in Section 3.6. 149. To evaluate the frequency of lacZ mutants in E. coli follow the protocol previously described [29]. 151. Once primary HUFs have reached confluency (D6-7), label desired amount of Corning ® CellBIND ® 6-well plates with date, day and passage number (D6, P1). For simplicity, this protocol defines the plating day as D6. Assign each well a culture ID. Make sure to differ between culture ID of treated and untreated cultures (e.g., X-1 for treated; C-1 for control). CRITICAL STEP Each HIMA should consist of at least 48 treated cultures; the upper limit depends on the experience of the scientist. Additionally, each HIMA should include an appropriate number of controls. Initially, we recommend at least 36 controls, but fewer controls (e.g., [6][7][8][9][10][11][12] can be included in subsequent assays when performing consecutive HIMAs in the same laboratory and by the same operator. 152. Detach the cells as described in Section 3.3. 153. Count cells using a hemocytometer and seed cells at 150,000 cells/well in 2 mL medium per well into 6-well plates. 154. To ensure an even distribution of the cells, move the flasks back and forth before placing them in the incubator at 37 °C, 95% humidity, 5% CO2, and 3% O2.  Table 2 and as described in Section 3.3 once a culture has reached 80-90% confluency. CRITICAL STEP As each well is a separate culture, evaluate its confluency separately. It is normal that some cultures grow faster or slower than other cultures and enter or exit senescence at a different pace. Dip the Pasteur pipette into 70% ethanol between aspirating media from different cultures to avoid carry-over of one culture to another. 160. Move cells to an incubator set to 37 °C, 95% humidity, 5% CO2, and 20% O2 five days posttreatment (D11) to induce senescence crisis. CRITICAL STEP When the first clones are emerging, wait until they have populated most of the well (at least half) rather than a smaller fraction of it. It is always recommended to split cells at lower dilutions initially (e.g., 1:1.5) and gradually go higher once the culture can repopulate the well quickly (e.g., 1:4, then 1:10). It is better to split the culture the next day again rather than losing the culture due to too harsh splitting. Cells can now be cultured in standard 6-well plates. NOTE Individual clones visually appear different: they often differ in size and morphology. Some clones grow quickly while others grow very slowly. 165. Continue checking cultures daily under microscope. Passage the culture as described in Section 3.3 at ratios indicated in Table 2 when first clones are emerging. TROUBLESHOOTING If wells appear empty after splitting and cultures are not able to recover and repopulate the well again, it is possible that the splitting ratio was too high ( Table 2). The culture should thus be split at a lower ratio next time. If the culture has been split too harshly and cannot recover, the remainders can be split again at a lower ratio. Keep the remainders of the cultures during critical steps, i.e., when first clones emerge, because sometimes it can be difficult to estimate the growth of a culture. If unsure how to split a culture, it can be split into two wells at different ratios (e.g., 1:3 and 1:4, or 1:5 and 1:10). CRITICAL STEP Perform the Nutlin-3a counter-screen once the culture shows a clonal population of homogenous appearance and has resumed growth. Generally, wait until the culture is able to repopulate a 6-well dish within six days after being split at least 1:3. Most cultures can be split at much higher ratios (Table 2). Again, each culture should be evaluated separately and split at a too low rather than a too high ratio. First clones will emerge as early as 25 days post-treatment (D32+), while some will take up until 83 days post-treatment (D90) of the assay to be ready for screening. 166. Prepare two wells on a 6-well plate for each culture that is ready for the Nutlin-3a counterscreen. Label the plates as shown in Figure 6 (Day 1) and add 2 mL growth medium to each well. 167. Passage the culture as described in Section 3.3 to two wells of the prepared 6-well plate. Choose the splitting ratio based on the last passages so it can repopulate the well in six days. 168. The next day prepare the Nutlin-3a treatment solution in a 15 mL tube. 169. Label the plate as shown in Figure 6 (Day 2). Aspirate the medium of both culture wells. Add growth medium to the control well (-N) and the Nutlin-3a treatment solution to the other well (+N). 170. Place the plate back into the incubator and visually inspect the culture under the microscope after five days to determine the response of the culture towards Nutlin-3a ( Figure 6; Day 7). TROUBLESHOOTING If the control well is not full after 5 days of treatment during the Nutlin-3a counter-screen, it is possible that the splitting ratio was too high. Passage the culture once more and rescreen at a lower splitting ratio. If the outcome of the Nutlin-3a counter-screen is unclear after 5 days of treatment, it is possible that the splitting ratio was too low or it may be a mixed-response culture. Sometimes WT-TP53 cells can appear insensitive to Nutlin-3a when they are very full at the beginning of the treatment. Passage the culture once more, then rescreen. If still unsure about the outcome, expand the Nutlin-3a-treated and untreated well and sequence for TP53 mutations.

Expansion into Immortalised HUF Cell Lines-Time for Completion: ~2 Months (D32-90)
171. NOTE Expand all cultures with a resistant and mixed response towards Nutlin-3a. Also keep the respective Nutlin-3a treated culture. Sometimes with mixed cultures the Nutlin-3a-treated well will have to recover for ~3 days before it can be moved to a T25 flask. Also expand at least one WT culture to prepare western blot lysates from and include on the western blot gel. 172. At the end of the Nutlin-3a counter-screen label two T25 flasks with culture ID, passage number and date (e.g., 124/124N, P10, 5/10/18) and add 5 mL growth medium per flask. 173. Passage both the control (-N) and the Nutlin-3a-treated (+N) well at an appropriate ratio from the 6-well plate to the flask. Base the splitting ratio on the ratio used for the Nutlin-3a counterscreen or adjust if well was over-/under-populated at the end of the Nutlin-3a counter-screen. NOTE Take the different surface areas into account. 174. Once the T25 flask is full repeat step 173 but passage the cells into a T75 flask with 12 mL of growth medium. 175. Once the T75 flask is full, label one T75 and two T25 flasks and add fresh growth medium into each flask. Passage the culture to the new flasks. 176. When the next T75 flask is full, trypsinise the cells and use the cell solution to prepare a pellet and two frozen stocks (continue to step 178-Section 3.6). 177. Make lysates for western blotting from the two T25 flasks (continue to step 188-Section 3.7).

Cryopreservation of Immortalised HUF Cell Lines and Preparation of Cell Pellets for DNA Isolation-Time for Completion: 30 min
178. Following from step 176 (Section 3.5.5), prepare frozen stocks and pellet once the T75 flask is full. 179. For each T75 flask label two 15-mL tubes with the culture ID and "Tube 1" (for pellet) and "Tube 2" (for stock). CRITICAL STEP It is possible to prepare frozen stocks from up to four cultures at a time. Because the freezing medium contains a high percentage of DMSO it is important that cells are moved to -80 °C immediately to avoid any cellular damage. 180. Label two cryogenic vials with culture ID, date, passage and size of flask (e.g., X-124, 10/07/18, P15, ¼ T75). 181. Label one microcentrifuge tube with culture ID and date. 182. Add half of the cell solution from step 176 (Section 3.5.5) to each of the 15-mL tubes from step 179 and centrifuge at 320× g for 5 min at room temperature. 183. Prepare freezing medium. 184. For pellet: transfer pellet from "Tube 1" to a microcentrifuge tube with 600 µL PBS, mix and centrifuge at 2300× g for 3 min at room temperature. 185. Aspirate the PBS from the pellet and store at −20 °C until DNA isolation (continue step 199-Section 3.9). PAUSE STEP Cell pellets can be stored at −20 °C for several months until DNA isolation. 186. For frozen stock: Aspirate the supernatant from "Tube 2" and re-suspend the pellet in 2 mL freezing medium (for 2× 1 mL stocks) by gently pipetting up and down. 187. Add 1 mL to each of the prepared cryogenic vials and place in a Styrofoam box at −80 °C.
Transfer vials to liquid nitrogen tank within 1 week.  Table 5 (Section 5). PAUSE STEP Amplified products can be stored at 4 °C until the next day or for a few days at −20 °C before continuing to next section. 221. Clean-up the PCR products from step 209 using the peqGOLD Cycle-Pure Kit according to the manufacturer's instructions.

TP53 Sequencing and Analysis-Time for Completion: 2 Days
222. Perform Sanger dideoxy sequencing on each product from step 221 (Section 3.10) with sequencing primers shown in Table 6 (Section 5). Sanger sequencing is offered by various DNA service providers such as GENEWIZ (www.genewiz.com). 223. Once the sequencing has been performed you will be provided with an .ab1 file.
PAUSE STEP The analysis of the .ab1 files can be performed at any preferred time. 224. Open the .ab1 file with a suitable software (e.g., Chromas, Technelysium Pty Ltd., Australia) and examine the chromatogram visually ( Figure 7A,B). 225. Reverse and complement the sequence by pressing "reverse" for all exons except exon 8_9 ( Figure 7A). 226. Export the FASTA sequence of each exon using "Ctrl+B". 227. Open the Basic Local Alignment Search Tool for Nucleotides (BLASTN) from the National Center for Biotechnology Information (NCBI) (https://blast.ncbi.nlm.nih.gov/Blast.cgi). 228. Paste the FASTA sequence from step 226 into "Query Sequence" box ( Figure 7C). 229. Copy and paste the TP53-WT sequence from Table A2 (Appendix A) into the "Subject Sequence" box ( Figure 7C). 230. Click "Blast" and print the result ( Figure 7C). 231. Find the beginning and end of the exon. Identify the mutation by looking for missing bases, mismatched bases or insertions (Figure 8). 232. Check the mutation in the chromatogram visually to identify whether it is a homo-/hemi-or heterozygous mutation (Figure 9). 233. Use the mutation feature search tool of the IARC TP53 mutation database (http://p53.iarc.fr/TP53GeneVariations.aspx) to assess information about the specific mutation. 234. Continue analysis with the other clones.

Optimisation of Carcinogen Treatment Conditions for HIMA
Because the HIMA is a long and labour intensive assay, experimental treatment conditions should be optimised prior to initiating the HIMA. Agents of interest should be tested for their ability to cause DNA damage in HUFs, and conditions must be optimised (i.e., concentration and/or exposure time) to ensure that the level of damage is high enough to potentially induce mutations while most cells remain viable. Existing knowledge of the mechanism(s) of action of the tested carcinogen can be a guide to choosing the experimental treatment conditions to test. When assessing cytotoxicity, the first test of concentration should be based on concentrations used in other cell assays reported in the literature. Initially, it is recommended to determine cell survival at a variety of concentrations and time-points (usually up to 48 h). Typically, for experiments in a 96-well plate, as described in this protocol, eight concentrations of the carcinogen and the solvent control can be examined in six replicates ( Figure 10A). Many carcinogenic agents require metabolic activation into reactive intermediates to exert their DNA damaging effects [30] and thus may require a longer treatment period (up to 48 h). When testing reactive intermediates shorter treatment periods (6-24 h) should be considered. To ensure that cells remain viable after initial treatment (i.e., continue to divide, fix TP53 mutation and continue to proliferate) it can be helpful to assess cytotoxicity after an additional 24 h sampling period. Our laboratory commonly uses the crystal violet staining assay as it is inexpensive, quick and reliable. We have also used resazurin-based assays (e.g., Deep Blue Cell Viability, Biolegend) but have found that some cytotoxic carcinogens, such as BaP, can stimulate the metabolism of resazurin even while cells are dying, giving incorrect viability results (unpublished observation).
To counteract the potentially deleterious effects of DNA damage, cells have evolved multifaceted mechanisms-collectively termed DDR-to detect DNA lesions, signal their presence and promote DNA repair [31]. Evaluating the induction of DDR signalling markers can help to estimate the presence of DNA damage in HUFs. However, it is important to note that the induction of DDR is not necessarily indicative of the mutagenic potential of the agent studied [21]. To standardise the experimental approach when assessing the expression of DDR markers, concentrations that cause 60-80%, 40-60%, and 20-40% cytotoxicity are typically selected. Usually, we examine the induction of phosphorylation or expression of four DDR proteins: phospho-p53, p21, phospho-Chk1 and phospho-γ-H2ax ( Figure 10B). Only those concentrations that show a clear induction of several DDR proteins should be considered for the HIMA.
Initial DNA damage, persistence of the damage, as well as the mutagenic specificity of individual DNA adducts, all contribute to the mutagenic potency of the carcinogens tested [26]. For many mutagens, DNA adduct formation correlates well with mutation frequency in standard mutation bioassays [32,33]. Thus, not only can the formation of premutagenic DNA adducts be used as a measure of the ability of the HUFs to metabolically activate the test agent, but also the extent of DNA binding (i.e., DNA adduct levels) helps to demonstrate that under the selected experimental conditions (i.e., concentration tested and exposure time) sufficient DNA damage is present to induce mutagenicity ( Figure 10C). Usually the same concentrations as those selected for the assessment of DDR markers are tested for DNA adduct formation. Only those concentrations that show detectable levels of DNA adducts should be considered for the HIMA. Several methods (e.g., 32 P-postlabelling, mass spectrometry) are available to detect and quantify DNA adducts [26]. All methods available for DNA adduct detection have their strengths and weaknesses, so the method of choice has to be decided on a case-by-case basis considering the carcinogen to be studied, prior knowledge of the carcinogen-induced DNA adducts formed and the infrastructure available in the host laboratory [26]. Alternative approaches to assess DNA damage (e.g., comet assay) have been considered in HUFs previously [34]. Again, only those concentrations that show a clear induction of DNA damage should be considered for the HIMA.
Bulky DNA adducts can be removed by DNA repair (e.g., nucleotide excision repair (NER)) [35]. As described earlier, we developed a Hupki mouse strain harbouring a knockout allele for a critical NER component, Xpa (xeroderma pigmentosum complementation group A) [15]. HUFs deficient in Xpa are incapable of removing bulky DNA adducts from their genomes. Thus, utilising Xpa -/-HUFs instead of Xpa +/+ HUFs may increase the TP53 mutation frequency in the HIMA. However, as reported previously, the mutation frequency was not increased in Xpa −/− HUFs although an increased number of mutations was found on the transcribed strand in Xpa −/− HUFs [15].
As the HIMA takes several months to complete, using a short-term reporter gene mutation assay (in a manner that is not dependent on the selection within HUFs) is another approach to optimise experimental treatment conditions prior to initiation of the HIMA. Mutagenicity can be rapidly screened in HUFs using the lacZ system [29]. We previously created HUFs in which the pUR288 plasmid locus has also been integrated [13]. The pUR288 plasmid can be extracted from the genomic DNA of mutagen-treated HUFs following 1-3 days of proliferation and lacZ mutations identified by selection in E. coli host cells [29]. Usually the same concentrations as those selected for the assessment of DNA adduct formation are tested for lacZ mutagenicity ( Figure 10D). Only those concentrations that show measurable levels of mutagenicity above background should be considered for the HIMA. It is also noteworthy that using the lacZ reporter gene in MEFs it has been shown that growing MEFs at standard oxygen (i.e., 20%) conditions increases the lacZ mutant frequency threefold compared to MEFs cultured at 3% oxygen. These mutations were mostly G→T transversions, which is the signature mutation for 8-oxo-guanine and an indicator for oxidative damage to DNA. This mutation type was not observed in cultures kept at low oxygen [7], highlighting the importance of performing carcinogen treatment in the primary HUF at 3% oxygen as recommended in this protocol.

Cell Morphology during HIMA
During the different stages of the HIMA the morphology of the cells will change (Figure 11). While primary HUFs used at the beginning of the HIMA are characterised by a spindle shape, the cells during the senescence crisis are flattened and more elongated. The first immortalised clones can emerge as soon as 25 days post-treatment (D32+). Immortalised clones appear in various sizes and morphologies and grow at different rates. They can be evenly spread across the well, but some of them grow in patches.

Results of the Nutlin-3a Counter-Screen
Once a culture contains a fully clonal population the Nutlin-3a counter-screen is performed.
Possible outcomes of the screen are illustrated in the scheme shown in Figure 12.  After a 5-day treatment with 10 µM Nutlin-3a the morphology of the Nutlin-3a-treated culture is compared with the untreated culture and the response to Nutlin-3a is assessed (see Figure 13). A culture can respond in a sensitive manner, which generally indicates a TP53-WT status, or in a mixed/resistant way, which indicates TP53-mutant status. Nutlin-3a-sensitive cultures show enlarged and significantly flattened cells that look very similar to senescent HUFs cultured at atmospheric oxygen. The growth of the sensitive cultures is inhibited after Nutlin-3a treatment, i.e., the well treated with Nutlin-3a is not full after 5 days. It is very common for a sensitive culture to fully recover after 3-7 days. Additionally, very rarely some sensitive clones carry TP53 mutations [22]. Resistant cultures show the same morphology and a similar growth rate to untreated cultures. Resistant cultures can fully populate the well within 5 days. Mixed cultures show a mix of senescent, sensitive cells together with areas of the well where the resistant cells are growing. A mixed-response culture will not be full after the 5-day Nutlin-3a treatment.
To further investigate the impact of the respective TP53 mutation on the p53 signalling pathway expression and induction of p53 pathway proteins (e.g., p21, p53 and Mdm2) are analysed by western blotting. For this immortalised HUFs are treated with Nutlin-3a for 24 h. A typical response of five TP53-mutant TP53-WT clones is shown in Figure 14. While treatment with Nutlin-3a leads to the induction of p53 pathway proteins (p53, p21, Mdm2) in TP53-WT clones (X-6-10), the response in TP53-mutants can differ ( Figure 14). Most mutants exhibit very strong basal p53 expression, which is not induced upon treatment with Nutlin-3a (X-1-2), while some mutations impact the size of p53 (X-3). Typically, most TP53-mutants do not show induction of p53 pathway members p21 or Mdm2 after Nutlin-3a treatment, but some mutants retain this ability (X-4 and X-5). Additionally, apoptosis and senescence signalling in HUF clones has been investigated by others using western blotting, but this is usually not part of the standard TP53 mutation assay in HUFs [14,36].

Results of TP53 Mutation Analysis in HUFs
After amplifying exons 4-9 by PCR, an agarose gel should be run to ensure that the PCR was successful and that the amount of DNA is enough for Sanger sequencing. A DNA ladder is run alongside the samples, which helps to estimate the sizes of the PCR products and the DNA amount. The size and intensity of each PCR product is compared to the ladder, where the 500 bp band is equivalent to 100 ng DNA and the other bands are equivalent to 50 ng DNA. An example of a representative agarose gel is shown in Figure 15. Once the sequencing data are available and FASTA sequences have been analysed by alignment against the human TP53 reference sequence, variations (e.g., single base substitutions, deletions) are assessed using the mutation validation tool of the IARC TP53 mutation database. TP53 mutations can be classified as either homo-/hemizygous or heterozygous. Typical results and a way to display them are illustrated in Table 3. Features that should be listed in the table are: codon number where the mutation is located, mutation type, strand on which the mutation is harboured, WT-and MUT-codon bases, zygosity, coding change, and the functional activity as assessed by Kato et al. [37] in a yeast promotor transactivation assay. Based on this yeast assay the impact of the mutation can be categorised as functional, non-functional or partially functional.

Carcinogen-Induced TP53 Mutations in HUFs
Environmental carcinogens that have been examined using the HIMA are summarised in Table  4. In each case, a unique TP53 mutation pattern was generated in the HIMA, which differed from that found in control HUFs that had undergone spontaneous immortalisation ( Figure 16).    Table 4. n = number of TP53 mutant clones generated for each agent.
Depending on the carcinogen tested the frequency of TP53-mutants was 11-33% while untreated controls had a TP53 mutant frequency of 4-13%. Agents studied include BaP, which is often used as a model for polycyclic aromatic hydrocarbons (PAH) that are formed during the incomplete combustion of organic material and are ubiquitous in the environment [33,41]. After metabolic activation by cytochrome P450 enzymes BaP is converted into its reactive intermediate BPDE, which forms pre-mutagenic adducts with DNA (i.e., 10-(deoxyguanosin-N 2 -yl) -7,8,9-trihydroxy-7,8,9,10-tetrahydro-BaP or dG-N 2 -BPDE) [15,42]. Using the HIMA the TP53 mutant frequency after BaP treatment (1 µM for 4-6 days) was 23% [11] and the TP53 mutant frequency for BPDE (0.5 µM for 2 h) was 19% [15]. However, it is important to note that the TP53 mutant frequency in the spontaneously immortalised HUF clones differed considerable between the two studies. Whereas the TP53 mutant frequency after BaP treatment was only 1.8-fold higher than background, for BPDE the mutation frequency was 4.2-fold higher [11,15]. The predominant mutation type in both studies was G→T transversion, which is in line with the formation of pre-mutagenic dG-N 2 -BPDE adducts. 3-NBA, a nitro-PAH present in urban air pollution and diesel exhaust particles, had a TP53 mutant frequency of 18-21% [13,38]. In the HIMA, G→T transversion was the predominant mutation type induced by 3-NBA in TP53 which is consistent with the prevalence of 3-NBA-DNA adducts formed at guanine in DNA [13,43]. Exposure to ultraviolet light induced a TP53 mutant frequency of 25%, with C→T transitions at dipyrimidine sites being the predominant mutation type in TP53 [39], which is in line with the formation of DNA photoproducts (i.e., cyclobutane pyrimidine dimers and (6-4) pyrimidine-pyrimidone photoproducts) [44]. The TP53 mutation pattern induced by the herbal drug aristolochic acid (AA) was dominated by A→T transversions which is in accordance with the formation of pre-mutagenic DNA adducts at adenine in DNA [45,46]. The TP53 mutant frequency in AA-exposed HUFs was 21-33% [12,39,40]. Table 4 and Figure 16 the HIMA has been used successfully to analyse the mutagenic potential and the TP53 mutation pattern for a variety of environmental carcinogens. In several cases, the TP53 mutation pattern generated corresponded to the pattern found in human tumours where exposure to these agents has been documented: (a) UV light, which is linked to human skin cancer, leads to the preferential induction of C→T and CC→TT mutations; (b) BaP or its reactive metabolite BPDE, which are associated with tobacco smoke-induced lung cancer, predominantly induce G→T mutations; and (c) AA, which is linked to urothelial cancer, and leads to a prevalence of A→T mutations. These data have been recorded in the IARC TP53 Database (www.p53.iarc.fr) which provides the basis to exploit the HIMA to examine current hypotheses on the endogenous or exogenous factors responsible for mutations in human cancers.

As shown in
Using AA as an example, a high prevalence of A→T mutations in TP53 is found in urothelial tumours of AA-exposed patients suffering from Balkan endemic nephropathy (BEN) originating from Croatia, Serbia, Bosnia, and Romania [47] ( Figure 17A). The same TP53 mutation pattern (i.e., A→T mutations) is found in urothelial tumours of patients suffering from aristolochic acid nephropathy (AAN) in Taiwan [47] ( Figure 17B). The A→T mutation type is otherwise rare in urothelial tumours [46] ( Figure 17C). As shown in Figure 17D the TP53 mutation pattern induced by AA in immortalised HUF cell lines closely mimics the pattern seen in AA-induced urothelial tumours in humans [12,39,40]. Therefore, the HIMA has sufficient specificity to make it applicable to other environmental mutagens that putatively play a role in cancer aetiology. Figure 17. TP53 mutation patterns found in (A) urothelial tumours of AA-exposed patients suffering from BEN originating from Croatia, Serbia, Bosnia, and Romania [47], (B) urothelial tumours of patients suffering from AAN in Taiwan [47], (C) urothelial tumours not associated with AA exposure [47] and (D) immortalised HUFs treated with AA [12,39,40].
As has been mentioned, many environmental carcinogens induce hotspot mutations in TP53 and these are often cancer specific [24,43]. As shown in Figures 16 and 18 the predominant mutation type in immortalised HUFs treated with BaP or its reactive metabolite BPDE is G→T and induced at several mutational hotspots in the TP53 gene [11,15,25]. The mutation pattern found in human lung cancer from smokers is also characterised by G→T transversions and commonly found at hotspot codons 157, 158, 175, 245, 248, and 273. As shown in Figure 18B, these mutational hotspots frequently observed in smokers' lung tumours are less common in non-smokers' lung cancer. In BaP/BPDEtreated HUFs, codons 157, 158, 248, and 273 are also recurrent sites of mutation, with a significant proportion being G→T indicating that data collected in the HIMA are consistent with the hypothesis that PAHs such as BaP have a direct role in causing TP53 mutations in smokers' lung tumours [24].  [11,15,25] and the spectrum found in lung cancer of (B) smokers, (C) non-smokers and (D) all cancers in exons 4-9. Reference for human cancer mutation distribution acquired from IARC TP53 mutation database, R20 July 2019 [48]. Exclusions for lung cancer mutation distribution: radon, asbestos, mustard gas, and coal [49,50]. Mutation hotspots are indicated in grey.

Future Perspectives: Investigating Mutational Signatures by Whole Genome Sequencing in Immortalised HUFs after Carcinogen Exposure
Several recent publications have reported using the HIMA immortalisation assay in conjunction with massively parallel sequencing (next-generation sequencing), conducted at either the wholeexome or whole-genome levels. Proof-of-principle reports have revealed successful identification of genome-wide mutational signatures of AAI, BaP, MNNG, UV light class C (UVC), as well as of the ectopically expressed activation-induced cytidine deaminase (AID). These signatures correspond reasonably well with the mutational signatures of the same exposures extracted computationally from the human cancer genomics data repositories [51,52]. Such scaled-up HIMA assay involving various chemical exposures have revealed remarkable gene-level recurrence of non-silent mutations in the mouse orthologues of known human cancer genes, accumulating across dozens of clones, including mutual exclusivity of mutations in components interacting functionally within the same protein complex. Thus, HIMA-derived cell lines can be used to study numerous candidate cancer driver events can be studied with biological relevance for functional impact on cell survival and immortalisation [53]. Lastly, the HIMA coupled with next generation sequencing and DNA adduct analysis has yielded a novel mutational signature of glycidamide, the reactive metabolite of acrylamide present in human diet and in tobacco smoke. Tailored computational analyses of the vast pan-cancer data were then able to identify the glycidamide signature in an unexpectedly large number of cancer types [54]. Collectively, these studies exemplify the future potential of the HIMA assay for systematic screening of carcinogenic compounds for robust mutagenic effects and functional impact, and they signal new in-depth insights into aetiology and development of cancer while generating important mechanistic evidence for adopting relevant preventive measures.

•
Growth medium: Mix 500 mL high glucose DMEM with 50 mL fetal bovine serum and 5 mL penicillin/streptomycin solution (10,000 U/mL/10,000 µg/mL). Label the bottle with the preparation date and store at 4 °C for up to 2 weeks. CRITICAL STEP All culture work must be carried out in a Class II biosafety cabinet to avoid assay culture contamination.  Label eight microcentrifuge tubes with the primer name (4F,  4R, 5F, 6R, 7F, 7R, 8F, 9R) and concentration (10 µM). Pipette 180 µL nuclease-free water into each tube, add 20 µL of 100 µM primer and mix well by pipetting. Make aliquots which can be stored for several years at −20 °C for at least two thaw-freeze cycles. CRITICAL STEP Keep on ice when not in use. • Sequencing primer working solutions (10 µM): Label four microcentrifuge tubes with the primer name (4seq, 5_6seq, 7seq, 8_9seq) and concentration (10 µM). Pipette 180 µL nuclease-free water into each tube, add 20 µL of 100 µM primer and mix well by pipetting. Make aliquots which can be stored for several years at −20 °C for several months. • CRITICAL STEP Keep on ice when not in use.  Table 7 below adding 15% extra. Add components to respective microcentrifuge tubes and place back on ice until use.  Table 8 into a 50-mL tube. Store at room temperature. Add protease-/phosphatase inhibitors fresh to working solution (10 µL/mL). • Prepare albumin standards from 0-2 mg/mL as indicated in Table 9. 10X TBS stock solution: Add 121.14 g Tris base and 175.32 g NaCl to 1800 mL water and mix on magnetic stirrer. Once everything is dissolved, adjust pH to 7.4 by adding concentrated hydrochloric acid. Complete volume accordingly and store for several months at 4 °C. • 1X TBS with 0.1% Tween-20 (TBST): Mix 200 mL 10X TBS with 1800 mL water, then pipette 2 mL Tween-20 in. Tween-20 is highly viscous and takes some time to be removed from the tip and dissolved. Mix on magnetic stirrer. Prepare fresh on the day. • 3% milk in TBST: Weigh 6 g powdered milk and add 200 mL TBST. Put on magnetic stirrer until dissolved. Prepare on the day. • Primary antibody solution: Label 50 mL tubes with antibody name, date and dilution. Add volumes of 3% milk in TBST, primary antibody and sodium azide as shown in Table 10. Primary antibody solutions can be stored for several months at -20 °C. • Secondary antibody solution: Prepare secondary antibody solution in 40 mL 3% milk in TBST as shown in Table 11 prior to use and discard afterwards.