This reminds me of the time when I was trying to knock down a tumor suppressor and although the luciferase reporter assay showed significant KD with 10 different shRNA candidates, not one of them could knock the protein down by western. One possibility is that your gene has isoforms and while you knock one down, the other is compensating for the loss at the protein level. I checked and seems your gene indeed has two isoforms. In this case, design an shRNA that targets both isoforms for degradation.

The only problem with this hypothesis is if you are seeing functional characteristics associated with knockdown (like increased cell death/proliferation) but no change in protein expression. Are you sure that the effects you see are due to knockdown and not from toxicity? It doesn't make sense that you see a reduction in mRNA, no reduction in protein but still see a functional impact.

This seems to be an interesting observation to me. I am considering that the notorious Central Dogma, which has been developed by using Escherichia coli, may not be true in humans and rodents. It is noteworthy that Escherichia coli is a minor bacteria in human gut.

By the way, I have searched DNA-binding protein SATB1/Special AT-rich sequence-binding protein-1 in vain among my protein database (please see file; HepG2 fucoidan).

Liver tissue (pseudo-liver cancer) has AT-rich interactive domain-containing protein 1A/SWI/SNF complex protein P270/B120/ARID domain-containing protein 1A at 9.8 μg/mg of tissue protein.

HCC tissue (named as No.6) has DNA-binding protein NEFA/Nucleobindin-2 at 0.34 μg/mg of tissue protein.

Therefore, this rare protein may not be expressed. Protein synthesis in humans seems to occur in accordance with the idea of famous Prof.Dr. Hans Neurath (born in Wien, Austria) that protein synthesis is occuring by the reverse reaction of many proteases in humans. Thus, proteases to produce SATB1 may not be present in your cell-line.

Actually, I think you may be looking at a protein with a long half-life....If it gets predominantly recycled and not degraded then, it will last beyond 72 hours...Also, you note that you see gene knockdown at 72 hours as evidenced by mRNA reduction...Maybe try to assess protein levels beyond 120 hours...It is not clear to me why cells die at 72 hours except that there is toxicity...

It would be of value to monitor protein degradation by the use of cycloheximide which blocks translation...

Let me know if you need more details about any of the suggested procedures...

This reminds me of the time when I was trying to knock down a tumor suppressor and although the luciferase reporter assay showed significant KD with 10 different shRNA candidates, not one of them could knock the protein down by western. One possibility is that your gene has isoforms and while you knock one down, the other is compensating for the loss at the protein level. I checked and seems your gene indeed has two isoforms. In this case, design an shRNA that targets both isoforms for degradation.

The only problem with this hypothesis is if you are seeing functional characteristics associated with knockdown (like increased cell death/proliferation) but no change in protein expression. Are you sure that the effects you see are due to knockdown and not from toxicity? It doesn't make sense that you see a reduction in mRNA, no reduction in protein but still see a functional impact.

Thank you all very much for your inputs and suggestions.

Regarding toxicity, I am sure the cells are not dying because of toxicity because I always have a control transfection with Luc3 siRNA where cells are not dying, rather growing faster than untreated ones.

For cells transfected with SATB1 siRNA, I see increased apoptosis, reduced mRNA, decrease in cell cycle rate but on protein level no difference.

The second isoform is SATB2. I blasted my siRNA sequence and confirmed it does not complement at all with SATB2. It may target other proteins like REEP1 or MAPK9 (84 and 73% match). The same siRNA was used by a previous PhD student who saw good knockdown in SATB1 even at protein levels (colon carcinoma cell lines).

Thanks for the clarification. Since you have a non specific siRNA and you don't see death there, I agree that your transfection protocol might not be inducing toxicity. However since you don't see a reduction in SATB1 protein, I would seriously doubt that the functional responses you note are from SATB1 knockdown. My thoughts are below.

1. Although an siRNA might work in one cell line/type, it may not be effective in others.

2. SATB2 is a protein that induced by a totally different gene. Both SATB1 and SATB2 seem to have two isoforms/splice variants (Uniprot). So you would have to blast your siRNA against the two splice variants of SATB1, to exclude the possibility of one compensating for the loss of the other.

3. At this point, I would also be looking at off-target effects. I would maybe start with the JNK pathway since you mention that your siRNA could target MAPK9. This pathway is involved in cell survival/proliferation. You could easily run a western blot to look that the expression of the JNK proteins.

I might also try to repeat the knockdown in the cell line the previous PhD student used, as a positive control.

Transcription repressor may be interesting, and I further searched of it. PDMD method is also applicable to human serum (please see file; JMBT Alopecia).

Fetal hepatocyte Hc has Coiled-coil and C2 domain-containing protein 1B/Freud-2 at 2.9, RecQ-mediated genome instability protein 2/BLM-associated protein of 18 kDa/OB DNA-binding domain-containing protein C16orf175 at 7.6 μg/mg cell protein, respectively.

Hepatoma HepG2 does not have it.

Healed hepatocyte HepG2 (cultured with fucoidan) has Teashirt homolog 1 at 0.13, Transcription repressor REST corepressor 3 at 2.6, Zinc finger protein 366/Dendritic cell-specific transcript protein at 2.3 μg/mg cell protein, respectively.

Normal liver (pseudo-liver cancer) has Zinc finger E-box-binding homeobox 1/NIL-2-A zinc finger protein at 1.6 μg/mg tissue protein.

LC tissue (with leprosy) has Transducin-like enhancer protein 1/Enhancer of split groucho-like protein 1 at 5.0 μg/mg tissue protein.

HCC tissue (with PBC) has Homeobox protein MSX-1/Hox-7 at 1.1 μg/mg tissue protein.

LC tissue (named as No.6) has Polycomb group RING finger protein 2/DNA-binding protein MEL-18 at 0.36 μg/mg tissue protein.

Serum of 8mo girl (with gastritis) has Atrophin-1-like protein/Arginine-glutamic acid dipeptide repeats protein at 9.4 μg/mg serum protein.

Serum of 1y girl (with biotin deficiency) has Zinc finger protein 674 at 2.0, and Scm-like with four MBT domains protein 2 at 12.0 μg/mg serum protein, respectively.

Healthy serum of 1y4mo girl has Zinc finger protein 263 at 4.7, and T-box transcription factor TBX3 at 4.0 μg/mg serum protein, respectively.

Serum of 4mo boy (with biotin deficiency) has PHD finger protein 12 isoform 1 at 3.6 μg/mg serum protein.

Healthy serum of the same boy as above (7mo1w) has Breast cancer-associated antigen BRCAA1/ARID domain-containing protein 4B/AT-rich interactive domain-containing protein 4B at 4.1 μg/mg serum protein.