Good catch! That (Halloran et al. 2012) reference should’ve pointed to (Sarbassov et al. 2006). I grabbed the wrong citation. I would edit the original post, but it’s now outside the editing window.
Can you say more about what concerns you here?
My overall intent in that post was to show that we’re not whistling in the dark. “We know enough to design more and less rational dosing schedules.” I focused primarily on avoiding mTORC2 inhibition, but the paragraph you quoted was about a second failure mode we want to avoid—the mTOR rebound. I closed out that paragraph saying: “I suspect this increase in Akt underlies the mTOR rebound sometimes seen with rapamycin, but that’s a topic for another post.” I wrote about it about a month later here: Rapamycin / MTOR Rebound effect in 3/12 non-GF and non-Keto patients – #62 by McAlister
As for the comparison, (Antoch et al. 2020) dosed Rapatar (the high-bioavailability formulation) at ~0.1 mg/kg and ~0.5 mg/kg. It’s worth noting that Rapatar’s bioavailability is only enhanced over unformulated rapamycin:
a single oral administration of Rapatar resulted in 12% bioavailability, which is comparable with commercially available formulations used in clinical practice (Comas et al. 2012)
The FDA label for Rapamune puts the bioavailability for tablets around 17.8% (27% higher than the oral solution’s 14% bioavailability). That makes Rapatar comparable to studies that deliver Rapamune through oral gavage.
I excluded these specifics in the original post because I didn’t think they were relevant to the conclusion 
I should also point out that I was comparing the oral Rapatar data with intraperitoneal data, not “with data from standard [formulations]”. The point was to show that high doses inhibit mTORC2, but lower doses (below the threshold of mTORC2 inhibition) can potentially lead to an mTOR rebound.
Sarbassov (10 mg/kg) and Schreiber (8 mg/kg) both used intraperitoneal injection, which results in much higher blood levels of rapamycin. Here’s a relevant figure from (Johnson et al. 2015):
Converting PPM to mg/kg (assuming 4g of food consumption and an average 25g mouse) gives us this:
| PPM | mg/kg BW (Approx) | Mean Rapa (ng/mL) |
|---|---|---|
| 14 | 2.24 | 12.8 |
| 42 | 6.72 | 124.8 |
| 126 | 20.16 | 402.2 |
| 378 | 60.48 | 709.8 |
Intraperitoneal injection resulted in about an 11-fold increase in rapamycin over a comparable oral dose of encapsulated, food-based Rapamycin (eRapa). The company claims “eRapa consistently provides approximately 30% more drug than generic rapamycin”, but that specific claim is based on unpublished company data.
TLDR: Rapatar’s bioavailability is comparable to Rapamune (and probably eRapa); my conclusion didn’t rely on similar bioavailabilities; I was comparing low and high doses to indicate that inhibiting mTORC2 was not the only failure mode we should try to avoid; confused at what you find problematically mixed 
Happy to nix “direct” from those sentences. Not sure why I included it, to be honest 
It makes no difference to my conclusion… which was that our dosing regimens should try to avoid both inhibiting mTORC2 and instigating an mTOR rebound.
This is incorrect. Rapamycin eventually inhibits the assembly of mTORC2 (Schreiber et al. 2015). I’m not aware of evidence that rapamycin reduces mTOR synthesis. Recycling a quote from my original post:
chronic exposure to rapamycin, while not affecting pre-existing mTORC2, promotes rapamycin inhibition of free mTOR molecules, thus inhibiting the formation of new mTORC2 (Sarbassov et al. 2006)
As with most things in biology, the mechanism is actually more complex. For example, in human rhabdomyosarcoma Rh30 cells, rapamycin inhibited the phosphorylation of mSIN1 (an mTORC2 component) within two hours and at very low concentrations (0.05 ng/ml) (Luo et al. 2015). This is distinct from rapamycin sequestering free mTOR.
Luo’s results built on (Rosner and Hengstschläger 2008), who showed that rapamycin triggered the dephosphorylation of rictor and mSIN1 (both mTORC2 components), followed by a translocation from the nucleus to the cytoplasm. This dephosphorylation and translocation resulted in a cytoplasmic rictor/mSIN1.1 complex that is not bound to mTOR. The end result is less mTORC2 assembly.