9. Enhanced non-homologous end joining contributes toward synthetic lethality of pathological RAD51C mutants with poly (ADP ribose) polymerase

Somyajit and et al., 2015. Carcinogenesis.

RAD51C germline mutations can result in Fanconi anemia-like disorder and make cells more susceptible to breast and ovarian cancers. These mutants were found to be hypomorphic in their functions.
Nuclear protein Poly (ADP-ribose) polymerase (PARP1) plays important roles in SSB repair and restart of stalled replication forks.  PARP1 inhibition can cause DSB accumulation by increasing amount of non-repaired SSBs and collapsed replication forks. Using specific inhibitors to target PARP in order to cause "synthetic lethality" in BRCA1/2-deficient cells in breast and ovarian cancers has been investigated in clinical trials. It was known that RAD51C deficient cells can be targeted by PARPi, but targeting hypomorphic RAD51C mutant cells by PARPi can be difficult.

Chinese hamster lung fibroblast cells:
CL-V4B (RAD51C -/-)
V79B (Parental)

PHEN (PARPi)
4-ANI (PARPi)
Camptothecins (CPT) activate S or G(2)-M arrest and the homologous recombination (HR) repair pathway in tumor cells. 

 "RAD51CC-deficient cells can be targeted by PARPi." CL-V4B cell survival was significantly decreased in dose-dependent manner (compared to wtV79B cells) after CL-V4b cells were treated with different 4-ANI doses (0.1-10 uM). Similar results were obtained in p53-defective HeLa cells and p53-proficient MCF7 cells. PHEN was demonstrated to generate the same effect in these cell lines. It was concluded that RAD51C-deficient cells are extremely sensitive to PARPi due to accumulation of CL-V4b cells in G2/M boundary and increased chromatid aberrations and radial chromosomes.

"Cells expressing RAD51C pathological mutants are sensitive to PARP inhibitors." Missense mutations RAD51C L138F, G125V, L219S and Q143R resulted significant reduction in HR activity, but RAD51C G264S, R366Q, D159N and T287A only show their hypomorphic function in HR. All of these 8 mutants showed sensitivity to CPT. To show if they were also sensitive to PARPi, CL-V4B cells were treated with 0.5 uM and 10 uM 4-ANI.  RAD51C L138F, G125V, L219S and Q143R were found to be highly sensitive to 4-ANI and show high levels of G2/M accumulation. But, RAD51C G264S, R366Q, D159N and T287A did not show much to the sensitivity and did not show much of the G2/M accumulation.

"PARP is hyperactivated in RAD51C-deficient cells due to replicative stress." PARP1 levels were similar in CL-V4B and wtV79B, but PAR formation controlled by PARP was significantly increased in CL-V4B cells. This hyperactive PARP activity was reduced with PARPi. PARP activity was high in RAD51C L138F, G125V, L219S and Q143R compared to in RAD51C G264S, R366Q, D159N and T287A. To test if the increase in PARP activity was due to replicative stress, they treated CL-V4B cells with hydroxuyrea to induce replicative stress. Interestingly, PARP activity was significantly elevated in these cells compared to the control. All of these results combine, they concluded that deficiency of RAD51C causes replicative stress which elevates PARP activity. To see how PARPi can affect the the replication process, CL-4B cells were treated with PARPi 4-ANI. Accumulation of PARP1 on chromatin and formation of DSBs were observed. So I can conclude that PARPi can be used to reduce PARP activity in RAD51C-deficient cells in order to generate "synthetic death" in cancerous cells caused by RAD51C-deficiency. 

"Inhibition of PARP in RAD51C-deficient cells leads to enhanced recruitment of NHEJ factors." RAD51C deficient cells showed significant reduction of RAD51 foci formation. When V79B cells were treated with PARPi 4-ANI, RAD51 foci formation was increased compared to that in CL-V4B cells. 4ANI-treated RAD51C-deficient cells showed reduced RAD51 recruitment on chromatin but higher amount of NHEJ proteins, KU70, KU80, Ligase IV and Artemis, there. In addition, the NHEJ downstream protein, DNA-PKcs, was also activated in these treated CLV4B cells.
"PARP inhibition induces error-prone NHEJ in RAD51C-deficient cells and suppression of NHEJ rescues PARP inhibitor-induced toxicity in CL-V4B cells." The NHEJ efficiency in vivo was measure using GFP-based NHEJ reporter. Linearization with HindIII produces cohesive 4 bp overhangs. But digestion with I-SceI produces an inverted overhang that required nucleolytic end process by certain proteins before end joining/circularization.   In CL-V4B cells, HR pathway is non-functional, leading to NHEJ to be more dominant even though NHEJ is an error-prone pathway. RAD51C-XRCC2 double depletion led to increase NHEJ in PARPi-treated cells. However, depletion of KU70 increased survival of 4-ANI treated-RAD51-deficient cells and rescued CPT-induced damage in RAD51-deficient cells, but made RAD51C-depleted cells become more sensitive to IR damage. Similarly, inhibition of DNA-KPcs, also one of the proteins required for NHEJ activity, or  Ligase IV depletion also reduced the toxicity of PARPi in CL-V4B cells.

"Stimulation of NHEJ synergizes 4-ANI-induced toxicity in cells that express pathological RAD51C mutants."Activation of DNA-PKcs in CL-V4b cells was increased after treatment of IR and PARP more than each treatment alone. They also tested the combined effect of Zeocin and PARP can affect NHEJ pathway with and without inhibition of DNA-PKcs in both V79B and CL-V4B cells. They found that Zeocin and 4-ANI simultaneously resulted in moderate increase in cells V79B and massive increase in CL-V4B cells in radial chromosome structures. However, these effects were rescued when DNA-PKcs was present. These results indicated that "synergersm"/combined effects in induction of genomic instability was by activating the NHEJ pathway.

Interestingly, CL-V4B cells that express some RAD51C mutants were resistant to PARPi even at high drug concentrations; also, low dose of IR(2.5 Gy) did not affect cell survival at all. Therefore, they tested if the combination of low dose IR and PARPi can change this result. As expected, this combination generated 10 fold cell death in RAD51C G264S, R366Q, D159N and T287A. [Why did not they test the cell survival in RAD51C L138F, G125V, L219S and Q143R?] When they overexpress these RAD51C missense mutants in U2OS-SCR24 cells, then treated these cells with IR and 4-ANI in combination, they observed that HR activity in RAD51C L138F, G125V, L219S and Q143R was reduced by 3 folds and HR activity in RAD51C G264S, R366Q, D159N and T287A was moderately reduced. Similar results of combined effects of IR and 4-ANI were obtained when using these mutants in HeLa cells. Notably, the levels of RAD51C expression in these mutants was compatible to wtRAD51C.  These results helped them conclude that RAD51C mutants can be treated with PARPi to cause DSB accumulation, and this effect can be further induced by using low-dosed IR and PARPi in combination to force cells to undergo NHEJ error-prone pathway, which leads to synthetic genomic instability and cell death.

8. Germline mutations in breast and ovarian cancer pedigrees establish RAD51C as human cancer susceptibility gene

Meindl A., et al., 2010

BRCA1, BRCA2, ATM, CHEK2, BRIP1 and PALB2 have been known to play crucial roles in genomic stability and homologous recombination in DNA repair, and disruption in any of these proteins can lead to breast cancer. Specifically, biallelic mutations in BRCA2 (FANCD1), BRIP1 and PALB2 (FANCN) lead to Fanconi anemia (FA). Recently, RAD51C R258H, a homologous missense mutation in RAD51C, was found to cause cells to be hypersensitive to DNA cross-linking agents and decrease RAD51 foci formation, which are typical features of FA. In this paper, they were interested in testing whether monoallelic germline variants in RAD51C can make cells more susceptible to breast and ovarian cancer, by screening more than 1000 patients who had gynecological cancers but negative for BRCA1 and BRCA2 mutations.

They found 14 RAD51C variants in 1100 affected, unrelated women who carry hereditary breast cancer and BC/OC. These 14 variants were catergorized into: 2 single-base pair insertions, 2 splice-site mutations, and 10 missense mutations. They also identified three missense mutations, A126T, G264S and T287A, which appeared to be more frequent in affect individuals from normal German controls.

A splice mutation 145+1G>T disrupted the canonical GT dinucleotide. It reduced the expression of functional RAD51C-001 and increased expression of nonfunctional RAD51C-008 while not affecting expression of nonfunctional RAD51C-009 compared to WT.

They tested 10 missense amino acids variants to see if they alter the function of RAD51C protein. They insert human RAD51C mutants into chicken DT40 cells where RAD51C ortholog was disrupted.
     - The survived cells expressed G125V and L138F at the highly conserved sequence, so they failed to complement the Rad51c mutant phenotype in DT40 cells.
     - RAD51C G3R, A126T, V169A and G264V complementary DNA corrected the mytomycin (MMC) hypersensitivity of Rad51c mutant in DT40 cells to the WT RAD51C cDNA levels.
     - D159N, G264S, T287A and R366Q partially restored the MMC sensitivity of Rad51c mutant in DT40 cells to WT RAD51C levels.
==> G125V and L138F did not restore normal RAD51 filament formation, while the other eight expressed the RAD51 foci formation just as seen in WT RAD51C.

Six germline variants were considered pathogenic, including two insertions, two splice site mutations, and two missense mutations G125V and L138F. However, two recurrent missense mutations, G264S and T287A were found to be not associated with cancer in 1100 tested women even though they resulted in "reduced cell survival and normal RAD51 foci formation,"

This paper established the first link between RAD51C and human cancers, breast and ovarian cancers, and confirmed RAD51C as a FA-gene, for the first time. It was suggested that women who who have strong history of breast and ovarian cancer, but are negative for BRCA1, BRCA2, and RAD51 mutations, should be screened using exomic sequencing.

First Week of PhD Rotation

Here I am. At the University of Pittsburgh, School of Medicine. Everyone is proud of me, and I think I am proud of myself, too. I still cannot believe that I am in graduate school now earning my doctorate in biology research two months right after being conferred my bachelor. I never imagined I would pursue this path when I arrived to the U.S. But since I joined research in the end of my freshman year, I have set my mind that this will be my forever interest. I always think that no ways I can get bored of research because there is a tremendous amount of unknown knowledge waiting to be discovered.  I had been waiting with all the excitement and pride for the moment of starting graduate school since my sophomore year. It is an amazing feeling to be accepted into a graduate program at a decent university.

After three days being in a new lab for my first rotation, I was amazed by everything; what people do, how they do it and all the available equipment. I realized that my undergraduate training is very weak and there will be so many things for me to fix. I was very disappointed that I was in lab for three years and now I have to relearn even the smallest things, such as where to dispose biological hazards, how to discard wastes properly and even how to open a tube properly, etc. to avoid all of potential contamination. I felt so ashamed that I did not ever pay enough to those little things before because no one my old lab does so. It was finally weekend, I went home with an exhaustion after long workday, but tried to read all the papers I was suggested and wrote summaries for them. First time working in lab for straight nine hours a day everyday and being all alone at home, having nothing to do but just focusing on research. I reviewed everything I did during those three days in the lab to make sure I would not make the same mistakes again. I got back to a good mood and set new goals for a new week, I would socialize better, stay more focused and avoid making the same mistakes.

New week began, it was a horrible week I believed because almost everything did not work out because of my bad techniques. I made new mistakes and sometime repeated my old mistakes. I was reminded several times on lots of things by the mentoring grad student. I was so annoyed and she was so annoyed that we always have awkward silence all the times. I was so angry at myself why I am so dumped and so not good at any techniques. I felt so so horrible wasting her time and materials. I really want to get better at what I am doing. I reviewed everything I have done throughout the day EVERYDAY, and I really want to be better, so I do not upset her and waste materials. But there are just so many things to remember and fix and erase from my old memories and replace my memories with new techniques. EVERYTHING comes at once! So.....I still have plenty of things to improve. 

Weekend again, I was totally down since I would not get any results, and I can feel her anger towards me. I am very upset about myself. I am sure she thinks I am very stupid and retarded, I cannot help myself in this case though because yeah...I am still adjusting to new things in this lab. I am trying my best but I am only 1.5 weeks into my graduate program. I am bad at my techniques, and this is why I need to be trained. I was from a very small university and was not trained by my mentor but by a master student. I could not help but carried out experiments the way I was taught. Now I am here with all awesome scientists, I see how experiments are properly handled. I am amazed and inspired! This is my starting point in graduate school and I hope she understands, I am trying my best to learn as quickly as I can and one day I will get to where she is now. I believe she still remembers how she felt when she was in my position now, even though it is years ago for her already.

--- I am encountered one of the hardest moments in my life......I hope I am strong enough to fight through this

7. Functionally null RAD51D missense mutation associates strongly with overian carcinoma

Rivera B., et al., June 23, 2017. Cancer Research.

RAD51S is an ATPase with the N terminal domian required for ssDNA binding. This domain includes a conserved ATP binding Walker A and Walker B motifs which are necessary for its role in HR. These motifs were also suggested to bind RAD51C and XCRR2 while residues 60-78 contribute to the  Linker domain responsible for interaction with RAD51C and XRCC2.

RAD51D c.620C>T; p.S207L (located in Walker B, highly conserved) has shown in this article for its role in disrupting the interaction of RAD51D and XRCC2 and increasing sensitive to PARPi. While the RAD51 paralogs, including RAD51B,C,D are known to involved in hereditary breast and ovarian cancer, but XRCC2 and XRCC3 are still not well understood. Truncated mutations of RAD51D in germ line cells have been known to predispose to ovarian cancer and cause cells to be susceptible to PARP inhibitors. Furthermore, high grade serous carcinoma (HGSC) is recently found to be associated with RAD51 missense mutations. Here, they found RAD51D c.620C>T; p.S207L mutation to be involved in ovarian cancer through studies of case-control, effect of this variant on the role of RAD51D in DNA repair activity and genomic profiling of known RAD51 tumors.

Control - RAD51D.3 cell line
Chinese Hamster Ovary (CHO) RAD51D deficient cell line - RAD51D.1
             stable expresses the c.620C>T; p.S207L - RAD51D.1-p.S207L.
 Wild-type RAD51D - RAD51D.1-wt

All RAD51D c.620C>T; p.S207L carriers do not contain BRCA1 and BRCA2 French Canadian founder mutations. However, they all have another missense RAD51D missense mutation, RAD51D c.698A>G; p.E233G. This result suggested that c.620C>T; p.S207L comes from chromosome carrying c.698A>G; p.E233G. [How? Why not c.620C>T; p.S207L causes the c.698A>G; p.E233G?]
The sequence data showed allelic imbalance increasing the ratio of RAD51D c.620T-mutant to c.620T-WT, ranging from 0.74 to 0.93, where 0.5 represents a balanced allele ratio. Loss of hereterozygousity at the RAD51D c.620C>T; p.S207L locus was found in all selected ovarian HGSC. TP53 somatic mutations were identified in all of them.  Since RAD51D c.620C>T; p.S207L located in Walker B, it disrupted the interaction between RAD51D and as seen in XRCC2  p.D206A mutation. RAD51D.1 and  RAD51D.1-p.S207L showed a significant decrease in RAD51 filament formation. Gene targeting frequency appeared to be significantly lower in RAD51D.p.S207L cells compared to RAD51D.1-wt, confirming HR defect.

 RAD51-p.E233G variant did not affect HR-dependent DSB repair; however, this mutation can rstore DNA repair activty in RAD51D deficient cells in the presence of p53, suggesting that pE233G does not negatively affect RAD51D HR activity by itself.


[Loss of heterozygosity (LOH) is a gross chromosomal event that results in loss of the entire gene and the surrounding chromosomal region.^[1]

Most diploid cells, for example human somatic cells, contain two copies of the genome, one from each parent (chromosome pair); each copy contains approximately 3 billion bases (adenine (A), guanine (G), cytosine (C) or thymine (T)). For the majority of positions in the genome the base present is consistent between individuals, however a small percentage may contain different bases (usually one of two; for instance, ‘A’ or ‘G’) and these positions are called ‘single nucleotide polymorphisms’ or ‘SNPs’. When the genomic copies derived from each parent have different bases for these polymorphic regions (SNPs) the region is said to be heterozygous. Most of the chromosomes within somatic cells of individuals are paired, allowing for SNP locations to be potentially heterozygous. However, one parental copy of a region can sometimes be lost, which results in the region having just one copy. The single copy cannot be heterozygous at SNP locations and therefore the region shows loss of heterozygosity (LOH). Loss of heterozygosity due to loss of one parental copy in a region is also called hemizygosity in that region.] - Wiki

PARPi causes RAD51D deficient cells by causing syntehtic lethality. PARPi (olaparib and talazoparib) or ciplatin impair the cell survival in cells expressing RAD51D-c.620C>T; p.S207L, similar to what was seenn in RAD51D.1 (RAD51-null) cells. This result helps confirms that PARPi can be a candidiate therapy for treating RAD51D c.620C>T; p.S207L

6. PARP Inhibitors in Clinical Use Induce Genomic Instability in Normal Human Cells

Ito S., et al., 2016.

Poly(ADP-ribose) polymerase 1  (PARP1) and PARP2 are quickly activated in response to DSB to produce PAR. Mice lacking PARP1 or PARP2 are hypersensitive to IR and alkylating agents and express genomic instability with increased sister chromatid exchanges (SCEs). "Neither PARP1 or PARP2 alone can compensate completely for the loss of other." PARP inhibitors (PARPi) were known to increase SCEs 2 folds higher in normal human cells at high doses. PARPi monotherary has been use to treat cancer in patients lacking BRCA1 and BRCA2 function where DNA repair is impaired leading to tumor reduction. Inhibiting nuclear PAR in proliferating cells by PARPi can result in DNA damage accumulation even in repair-proficient cells and cause genotocixity. An increase in SCEs can potent mutants and induce loss of heterozygosity.

The experiments were done on non-transformed cells (MCF-10, HMEC-hTERT) under the treatment of olaparib, which is an effective inhibitor of PARP1 and PARP2. The result show a significant elevation of SCE frequency in normal proliferating cells in response to 1uM olaparib, leading to genomic instability from hyperrecombination.

It was hypothesize that highly induced SCEs only occurs when PARPi is present, which is contrary with the known fact that DNA damaging agents usually result in persistent effect in structural DNA alteration by increasing SCEs.  Therefore, after exposure to 1uM olaprarib for 24 hours, it was demonstrated that olaparib exposure greatly induce chromatid-type aberrations, decreasing genomic integrity.

SCEs is also dose-dependent to olaparib and veliparib with doses even below IC10 . The MCF10A was able to keep the survival up to 95% but elevated SCEs up to 4 folds. SCEs are elevated at low doses of olaparib and veliparib but also further induced without increasing cytotoxicity as residual PARP activity is inhibited.

BRCA1 and BRCA2 deficient cells are highly sensitive to olaparib and veliparib in mouse emryonic stem (mES) cells. DNA damage was induced with chemotherary while inhibiting SSB repair with PARPi. Cells were treated with cisplatin and olaparib to compared their combined effect to that of individual agent alone. They showed additive effect in increasing the SCEs to 1.8-1.9 folds in MCF10A (even with non-cytotoxic concentration of cisplatin), potentially leading to chromatin-type aberrations.

Two cytogenetic biomarkers, SCE and chromatid-type aberrations were analyzed to confirm that continuous exposure to PARPi, especially when combining with cisplatin, can cause genomic instability in proliferating cells. The increased chromosome aberration should be carefully considered when using "these agents in prevention and early stage cancer and non-oncologic indications."

"The hyper recombination with elevated SCEs is agreeable with unresolved breaks persisting and disturbing normal DNA replication. An SCE is  the cell's "next best" recourse when presented with DNA damage at a stalled replication fork. Yet, the increase in chromatid aberrations confirms that the increase in SCEs does not fully succeed in rescuing all the DNA breaks occurring with PARP inhibition."

5. Novel insights into RAD51 activity and regulation during homologous recombination and DNA replication

Stephen K. Godin, Meghan R. Sullivan, Kara A. Bernstein. Review paper. 2016. 

IR induces DSB
MMS induces replicative stress

This paper focuses on understanding the homologous recombination under Rad51 regulation. Rad51 filament formation/presynaptic filament/nucleoprotein formation is well regulated to avoid unnecessary recombination and genomic instability.

In budding yeast, the Rad51 regulators include Rad51 mediators, scRad52, 2xRad51 paralogues, scRad55 and scRAd57. In budding yeast (sc), Rad51 regulators help to overcome the barrier by replication protein A (RPA) coating the 3' overhanging ssDNA after the resection. However the RPA removal mechanism is not clearly understood.

Human Rad51 paralogues form 3 complexes. The BCDX2 complex is composed of hRAD52B, hRAD51C, hRAD51D and hXRCC2. The XC3 complex is composed of hRAD51C and hXRCC3. Shu complex is composed of the hSWSAP1 (one of the RAD51 paralogues)  and the hSWS1.

scRAD52 is analogous to hBRAC2 in their sequences so it plays similar role in RAD51 foci formation, but hRAD52 also exists in human

scRAD51 containing Walker A and Walker B ATP binding motifs that only bind ATP-bound ssDNA. There are other anti-recombinase proteins that prevent the illegitimate HR to ensure that RAD51 filament is only formed at the DNAdamage sites in order to avoid excess recombination and DNA instability.

When the scRAD51 filament is established, the strand invasion can start on dsDNA to search for homologous stretch of DNA. Once scRad51 filament passes 15-nucleotide threshold, it stops sampling another dsDNA and becomes committed to the replication based on the target dsDNA. Then, scRAD51 was dissembled by scRAD54 to allow the invasion of DNA polymerase to the heteroduplex to extend the invading ssDNA up to 1700 bp. In human, hRAD54 was also known to direct hRAD51 invasion and disassemble hRAD51 from the ssDNA to allow DNA pol to start its job.

In the simplest form of repair, synthesis dependent strand annealing (SDSA) or double Holliday junction (DHJ) or break-induced replication (BIR) can take over the role to produce cross-over or non-cross-over products as represented in the diagram.

DNA lesions can initiate homologous recombination at the replication fork by stalling the DNA or inhibiting the progression of the replication fork. Each lesion is repaired by a different mechanism. Her we focus on HR at stalled replication fork. DNA alkylating agent methylmethane sulfonate (MMS) can result in DNA lesion that requires HR to DNA repair. During HR, a X-structure was formed by scRAD51 and accumulates in the absence of scSGS1. Interstingly, HR in MMS-induced DNA lesion occurs at gapped ssDNA but not free ssDNA ends. 

  

hRAD51C and hRAD51 were considered as Fanconi anemia but truncated hXRCC2 was also detected in this disorder. 

Mouse KO models have resulted in embryonic lethality, making our understanding of mammalian RAD51 paralogues limited to specific cell lines. Disruption in any RAD51 paralogues can result in chromosomal instabiloty, increases in MMC sensitivity, and decreases in cell survival.

In human cell lines, osteosarcoma U2OS and breast cancer MCF7, the BCDX2 and CX3 perform different functions at different stages in HR. BCDX2 is important for upstream of hRAD51 filament formation while CX3 is essential for downstream hRAD51 foci formation in response to IR. Depletion of either BCDX2 or CX3 complex caused a decease in HR at an Depletion of both BCDX2 and CX3 complexes did not yield additive effect on levels of hRAD51 filament formation.

Recruitment of hBRCA2 is independent of both BCDX2 and CX3 during IR-induced HR (because PALB2 and BRCA2 forms a complex at the damage sites (Park et al., 2014)), BUT hBRCA2 is epistatic (needed) for RAD51 paralogues to accumulate at the DSB. RAD51 paralogues are not epistatic with hRAD52, which performs a different function from BRCA2 does. Depletion of hRAD52 with hXRCC3 caused synthetic lethality and decreased HR rates, saying that hRAD52 and hRAD51 paralogues function in different pathways [So hRAD52 does not promote HR which is RAD51 dependent????? When does hBRCA2 and hRAD52 compete in repairing damaged DNA????]

RAD51 paralogues respond to damaged replication fork by performing these functions:
     "1) Protect nascent ssDNA at stalled forks from resection
       2) Prevent replication fork collapse 
       3) Facilitate replication restart"

Shu complex is an obligate heterotetramer comprised of scShu2, scShu1, scCsm2, and scPsy3. Disruption of any Shu components will result in increase sensitivity in replication fork stalling. The Shu complex is likely to act upstream of hRAD54, suggesting that this complex may play a role in hRAD51 filament formation. The scCsm2 and scPsy3 were crystalized and found to mediate the Shu complex in binding DNA and are structural scRad51 even though it does not share much sequence homology with scRAD51. The Shu complex interacts with the scRad1 filament machinery via Rad55-Rad57 bridge in presence of scRad52. Since scSrs2 interacts with the ATPase domain of scRad51, causing ARRP to hydrolyze, leading to the inability of scRad51 to bind to DNA,  scRad55-Rad57 can directly prevent scSrs2 from destablizing Rad51 filament in vitro. These results present a direct role for the Shu complex in scRad51 filament formation. 

"hRAD51C and hXRCC3 were found to assemble into an additional complex that is scaffolded by hPALP2, termed the "HR complex" comprised of hPALP2, hRAD51C, hRAD51, hBRCA2 and hXRCC3 (Park et al., 2014)." Heterozygous mutations in hPALB2, hBRCA2, and hRAD51C can lead to hereditary breast and ovarian cancers because of the lost WT copy. Homozygous mutations in hPALP2 (FANCN), hBRCA2 (FANCD1), and hRAD51C (FANCO) can lead to FA condition. 

IR-induce DSB causes disruption in hRAD51 which will inhibit HR pathway and promotes NHEJ pathway. tumors with disrupted hRAD51 can be targeted by PARP inhibition. However, mutation in hRAD51 in combination with PARPi can result in collapsing replication fork and activation of error-prone NHEJ pathway which can increase cell death. Therefore, clinical practice need to consider if using PARPi is safe to treat breast and ovarian cancers. 

4. Structural Basis for Recruitment of BRCA2 by PALB2

Olivier A., et al., 2009. EMBO

PALB2 is known to regulate BRCA2 to DSB, and BRCA2 is essential for RAD51 recruitment to form nucleprotein filament.

PALB2 is known as FANC-N, Fanconi anaemia subtype N
                                 130  kDa
                                 No clear functional domain other than predicted amino-terminal coiled coil structure and a carboxy-terminal WD40 repeat motif.
                                 The WD40 forms a 7 blades on PALB2 C-terminus as a ring structure. Y1183X mutation loses some C-terminal residues that inhibits the closure of the WD40 ring and destabilize PALB2 protein.

PALB2 N-terminus was thought to bind to BRCA1 to help recognize the DSB. Then, The N-terminus of BRCA2 protein interacts with PALB2 at its bottom pocket formed by the 4th and 5th blades and also interacts with EMSY protein at exactly at the same sequence. [EMSY is highly amplified and expressed in breast and ovarian cancers]. So in cancerous cells, EMSY may outcompete PALB2 at the binding site on BRCA2 to disturb the HR-mediated DNA repair.

[How???? Does BRCA2 compete with BRCA1 when BRCA1-PALB2 recognize the DSB???? or they bind at different sites or they just bind PALB2 at different stages during HR???]. BRCA2 interacts with RAD51 via the BRC repeat in the BRCA2 C-terminus. [So BRCA2-PALB2 complex is formed at the DSB to attract RAD51 over bc BRCA2 is scaffolding protein, so how does BRCA2-PALB2 complex signal for RAD51 to come to damage sites????] This BRCA2 C-terminal binding domain can revert mutant BRCA2 protein's internal deletion to  restore HR. 

3. BREAST CANCER-ASSOCIATED MISSENSE MUTANTS OF THE PALB2 WD40 DOMAIN, WHICH DIRECTLY BINDS RAD51C, RAD51 AND BRCA2, DISRUPT DNA REPAIR

WD40 domain - 7 blades beta-propeller fold 

Park J.Y., et al., 2014. Oncogene.


BRCA2 mediates the oligomerization of the Rad51 recombinase that is needed to form nucleoprotein filament with ssDNA. PALB2 associates with BRCA1 through its N-terminal coiled coil domain and with BRCA2 through its C-terminal WD40 domain. WD40 domains have ring structure of beta-propeller with seven blades that regulate protein interactions.

3 missense mutations of PALB2 WD40 domain have been found to disturb PALB2 protein more than its truncated mutations do. The disruption of PALB2, RAD51C, and BRAC2 interactions leads to breast cancer, overian cancer, and Fanconi anemia.

HF-RAD51C is purified using His6-FLAG-RAD51C based on which interactions with BRCA2 and PALB2 were detected via Co-IP. RAD51C was found to bind to WD4 domain of PALB2. However, Rad 51D does not interact with PALP2-BRCA2-RAD51C.

They created three missense mutations in the WD40 of PALP2 protein, including L939W, T103I, and L1143P at three different binding surfaces of PALB2 DW40 beta-propeller structure. The result suggested that PALB2 WD40 may scaffold the complex of HR and FA protein, and each HR protein is thought to bind to DW40 domain in independent manner at each distinct surface. XRCC3 is directly bound to PALB2 that resembles that of BRCA2, but completely distinct from that of RAD51C. These results may mean that RAD51D and XRCC3 interact with PALB2 in independent manner.   Therefore, formation of PALB2 - RAD51C complex or PALB2 - XRCC3 complex is doubted to compete with formation of RAD51C-XRCC3 complex, which is promoted in the absence of PALB2. [Does it mean RAD51C directly bound to XRCC3 in RAD51 paralogs, but they are in the same subcomplex CX3 in RAD51 paralogs??? I think M is doing experiments on this] 
Interestingly, RAD51C binding to PALB2 seemed to release BRCA2, leading to an increased amount of BRCA2 protein in mutant forms, and promotes RAD51 recruitment.
Since PALB2 is needed for RAD51 foci formation and CX3 only affects downstream of RAD51 foci formation. Therefore, it can be hypothesized that BRCA2-PALB2 was established first then RAD51C binds to PALB2, kicking BRCA2 out to promote RAD51 nucleoprotein formation, then PALB2 leaves after RAD51 foci formation to let RAD51C interacts with XRCC3 to form CX3 subcomplex for downstream of RAD51 foci formation in HR activity. But when does PALB2 interact with XRCC3 when PALB2 is busy in BRCA2-PALB2 complex, then in PALB2-RAD51C complex before RAD51 foci and needs to leave after RAD51 foci to allow RAD51C-XRCC3 complex formation? So...what is the purpose of PALB2 to bind to BRCA2 ????? Based on Olivia, 2009, PALB2-BRCA1 recognizes DSB sites, but how can they give a signal to recruit RAD51C or RAD51 paralogs to come to the damage sites when we know BRCA2 is a scaffolding protein? And what distinguishes the roles of BRCA1 and BRCA2?

L1143P mutant significantly reduced RAD51 filament formation, which is consistent with a significant decrease in DSB-HR associated with this mutant. L939W and L1143P also significantly increase in sensitivity to the IR.

In the opposite side, mutants of RAD51C were also tested, including L138F and D159N (breast and ovarian cancers) and R258H (FA), for interaction with PALP2. These mutants showed a decrease in binding of RAD51C to BRCA2, especially to PALB2.

RAD51 seemed not to be necessary for RAD51C or BRCA2 to bind to PALB2.

Nonfunctional truncated (complete pr partial removal) of DW40 in PALB2 protein are found have completely no DSB-initiated HR activities, but PALB2 missense mutations mentioned above only partially disrupt DSB-initiated HR pathway.

As a result, L939W, T103I, and L1143P mutations in PALB2 WD40 lead to genomic stability by at least two pathways. First, T103 is unstable in its expression in human cells, decreases in T103 WD40 PALB2 in cells may diminish its function in HR. Second, L939W and L1143P WD40 PALB2 suggested that they may alter the binding of PALB2 to HR proteins in vitro and human cells.

PALB2, BRCA2, and RAD51 are related to inherited breast cancer, while upstream FA pathway proteins are not.

2. Rad51 Paralog Complexes BCDX2 and CX3 Act at Different Stages in the BRCA1-BRCA2-Dependent Homologous Recombination Pathway

Chun J., 2013

Terminology 
                      HR - Homologous recombination
                      IR - Ionizing radiation
                      gene conversion - "one DNA sequence replaces a homologous sequence such that the sequences become identical after the conversion event" https://en.wikipedia.org/wiki/Gene_conversion
                     Resolvase = recombinase = nuclease that involve in DNA recombination
                     Supress sgs1 hydroxyurea (HU) sensitivity = SHU complex.
Summary

BRAC2 is the main mediator to promote Rad51 filament formation. In the absense of BRCA2, Rad52 drives a secondary pathway for Rad51-dependent HR. Depletion of Rad52 and BRCA2 further reduced HR. Since depletion of any Rad51 paralog did not further reduce the HR in BRCA2-depleted cells, Rad51 paralogs are epistatic with BRCA2.
    => Depletion of BRCA2 causes more significant effects in HR than the depletion of any Rad51 paralog, most likely because BRCA2 holes a direct role in loading Rad51 onto the DSB and Rad51 paralogs are more likely to stabilizing and protecting the Rad51 nucleoprotein filament from  antirecombinase.
         However, Rad51 paralogs may hold more significant effects in repairing replication-stalling lesions since they have more hypersensitivity to replication-stalling agents (DNA cross linking agents or topoisomerase inhibitors) than IR.

Rad51 carries out the recombinase activities was well studied, and Rad51 paralogs are known to facilitate Rad51 in HR under the IR condition. However, the molecular mechanisms of Rad51 paralogs in this process (Rad51B, Rad51C, Rad51D, XRCC2, XRCC3) still needs further investigation.  So, Rad51 paralogs can be divided into 2 complexes, including BCDX2 (consisting of Rad51B, Rad51C, Rad51D, and XRCC2) and CX3 (consisting of Rad51C and XRCC3)

Rad51 paralogs also relate gene conversion tract length and Holiday junction recombinase activity. In mice, diruption of any Rad51 paralog can cause lethality in early embryonic stages and unrepaired DNA accumulation.

The data in this paper suggested that BCDX2 and CX3 affect the HR pathway at different stages, in which BCDX2 acts downstream of BRCA2 recruitment but upstream of Rad51 recruitment to the DSB and stabilizes the damage sites, while CX3 acts downstream of Rad51 filament formation.

Disruption in Rad51C significantly affects the CX3 complex but not BCDX2 complex. Depletion of any Rad51 paralog significantly reduces HR activity demonstrated by measurement of I-SceI-induced homology-mediated repair.

Both Rad51 complexes are necessary for downstream activity of BRCA1 or BRCA2 but before the Rad51 nucleprotein formation. The depletion of CX3 did not affect the Rad51filament formation. However, the depletion of both BCDX2 and CX3 led to a decrease in Rad51 filament formation. Therefore, BCDX2 acts upstream and CX3 acts downstream of the Rad51 recruitment.

Since depletion of any Rad51 paralog with Rad52 further decreased HR frequency measured by amount of I-SceI protein, demonstrating that Rad51 complexes a synthetically lethal relationship with Rad52.

In yeast, Rad51 paralogs (Rad55-Rad57 complex) were known to facilitate Rad51 nucleprotein formation. Yeast Rad55 ~ human XRCC2, yeast Rad57  ~ human Rad51D. The Shu complex prevents the disassembly of Rad51 by Srs2 from DNA.  The Shu complex consists of 4 subunits, including Csm2, Psy3, Shu1, and Shu2.
 
Yeast Shu1 ~ Human XRCC2
Yeast Psy3 ~ Human Rad51D
Yeast Shu2 ~ Human SWS1, which interacts with Rad51 and Rad51 paralogs through hSWS1-associate protein.

Since XRCC3 depletion significantly reduced HR activity but did not alter Rad51 filament formation, therefore, it was hypothesized that CX3 complex is not responsible for Rad51 recruitment to DSB. Depletion of Rad51C can significantly remove CX3 complex, causing more significant damage in the CX3 complex than in the BCDX2.

Need Clarification
- DHJ Holiday junction
- HR
- NHEJ
- SDSA
- Plating efficiency