[Babcock_Hall]

I am working out how to make one or more dipeptides in which a modified aspartate residue is at the C-terminal position.  I am tempted to start by making a side-chain Weinreb amide from a suitably protected derivative of aspartate (tertiary-butyl ester at the carboxylate group), as we have done before.  Then I would deprotect the nitrogen and form an amide bond to another protected amino acid.  I am considering buying or making N-hydroxysuccinimide derivatives of the N-terminal amino acids (I found a paper in which dipeptides were made in this manner).  Then I would reduce the Weinreb to an aldehyde and do additional chemistry to introduce an electrophile, and finally deprotect.

My first question is whether or not the N-hydroxysuccinimide choice is a good one.  My second question is whether or not I have chosen the optimal point in the synthesis to make the peptide bond.  My third question is whether it makes a difference to begin the synthesis with a derivative of aspartate that is BOC- or FMOC-protected (it would be protected at the carboxylate in either case).

[kriggy]

The choice of protecting group itself has impact on the reaction conditions you are going to use. Fmoc is base labile therefore you cant use strong bases with fmoc. Its cleaved by piperidine and other bases. AFAIK DIPEA and lutidine are fine because the deprotection with them is quite slow.
For Boc chemistry, you are limited to non-acidic reactions. I think both are fine for making the dipeptide you want but you need to consider the steps further down the line. Im not sure what are the conditions for reduction of Weinreb but you can actually reduce N-Boc to N-methyl with hydrides which is another side-reaction to consider.

Furthermore, since you are using tbutyl ester, you cant use Boc as a protecting group for amine because both are acid labile. While it is possible to deprotect Boc in presence of tBu ester, it seems easier to use Fmoc AA which is also detectable by UV/VIS.

I would use the tButyl ester of the Weinreb and Fmoc AA (be it OSU ester or OH with activating reagents)

[Babcock_Hall]

I can answer the third question myself, now that I have thought about it; it has to be FMOC.  In our previous work, we started with a N-BOC, tertiary-butyl protected amino acid and went through the Weinreb, to the aldehyde with Dibal-H, and then performed a Horner Wadsworth Emmons reaction to install our electrophile.  Then we deprotected the BOC and tertiary butyl groups together.  My present thinking about the peptide is to replace the protecting group on the C-terminal residue with a protected amino acid residue in two steps, then move forward as we did before.

[Babcock_Hall]

I have been reviewing the literature of making dipeptides containing vinyl sulfones.  We still need to choose a coupling reagent, and I am focusing on three right now, based on the reading that I have done so far.  One is ED(A)C/HObt; two is HBTU, and three is using a N-hydroxysuccinimide ester.  My question is how easy or difficult is it to extract the urea by-product of EDC using water.  I have seen one protocol* in which only water was used, yet I have also seen a protocol in which 1 M HCl was used.  The latter protocol did not have a BOC-protected intermediate, but our synthesis probably will.  I would like to stay away from an acid wash, if I could avoid it.
*this protocol was for the synthesis of a tert-butyl ester, not a peptide; therefore, it may not be relevant.

[rolnor]

I think dilute HOAc is sufficient to remove the EDC-byproduct.
I am worried about your aldehyde, you have amides/carbamates in this molecule, maybe its stable for short time, maybe not. Coupling with a hydroxysuccinimide-derivativeis very clean and fast, I think its the most convenient way if you can buy the activated aminoacid.
You are starting to get many functional groups in this molecule so heads up!

[Babcock_Hall]

Thanks for the tip about HOAc.  My present strategies are either to couple in the presence of a Weinreb amide or to couple in the presence of a vinyl sulfone, but I prefer not to couple in the presence of an aldehyde.

[rolnor]

Yes, that sounds logic. I dont know the stability of the Weinreb, the vinylsiulphone is a Michael-acceptor so it can be a problem with amine adding to this double-bond.

[wildfyr]

EDC byproduct is indeed readily water soluble.

NHS ester chemistry is very well defined and clean.

[Babcock_Hall]

Right now I am leaning toward the route in which an FMOC-protected nitrogen-containing compound is made into the aldehyde, then subjected to a Horner Wadsworth Emmons reaction to make a vinyl sulfone, then deprotected on nitrogen prior to coupling.  Based upon some preliminary reading, I became concerned that our usual Horner conditions, which use DBU as the base, would be unsuitable for the FMOC-protected compound.  I realize that one could use TEA or DIPEA, which don't deprotect FMOC-protected nitrogens as quickly as secondary amines, but I did a literature search last night to seek other alternatives.  I found very similar reactions in which lithium hexamethyldisilazane (LiHMDS) or NaH were used, and those two bases will be my first two choices.  However, one reaction to produce an acrylamide did use DBU.  I have not read that paper yet, but it occurs to me that if the removal of the proton is faster than the removal of the FMOC group, this might explain their success.
EDT
Besides NaH and LiHMDS, tertiary butoxide might work, based on Reactivity Chart 8 in Greene and Wuts.  There might be an advantage in lithium over potassium as the counterion, in order to chelated the Horner phosphonate.  I have not researched this yet, and I would not try it unless there were precedent, because I am not trained in synthetic method development.  I did use potassium tertiary-butoxide successfully one time in a Horner reaction, but it was a bit of an odd duck as reactions go.

[rolnor]

You are asking for trouble with strong base and Fmoc, if you can use TEA its a lot better.

[Babcock_Hall]

The papers which used LiHMDS or NaH had modest yields (in the ballpark of 30%), despite coming from labs with experienced synthetic chemists.  The fourth edition of Greene and Wuts present a table of half lives for removal of FMOC.  20% piperidine is 6 seconds and 50% DIPEA is 10.1 h.  I would suppose that DIPEA is slightly slower to remove FMOC than TEA, but TEA does not appear in their table on p. 713.

Given that I found one successful use of DBU to make an acrylamide (which is within shouting distance of a vinyl sulfone in terms of reactivity), I also wonder if the key to success is to consume the base more quickly in removing the proton from the Horner phosphonate than it removes a proton from FMOC.

[kriggy]

I dont recall any of my SPS friends mentioning any troubles with TEA or DIPEA in peptide couplings or any other SP reactions but maybe their reactions were quicker than 10hrs.

I wonder if you could solve the possible deprotection issue by first reacting the phosphonate with 1 eq of base such as LiHMDS and then, after deprotonation, add the aminoacid aldehyde. Ive run HWE only once and I  just mixed all my stuff toghether but pre-forming the ylide might work as well

[Babcock_Hall]

Kriggy,

The order of addition is a good point that I had not considered.  When we use DBU, we always add the base to the phosphonate.  I will look at the order of addition used by the authors using other bases in closely related reactions.

[rolnor]

Yes, I dont know the acidity of the phosphonate, maybe its more acidic then the Fmoc-proton. Its a good suggestion to mix the base and phosphonate first.

[Babcock_Hall]

I wanted to return to the issue of reduction and assume for the moment that we reduce at the stage of having a BOC-protected dipeptide-like compound (right now this is strategy #2).  I was reading about solid phase synthesis of peptidyl-aldehydes in Chapter 6 of the book Fmoc Solid Phase Peptide Synthesis, A Practical Approach.  Some groups attach the peptide via a Weinreb amide-like linker.  When they reduced with LiAlH₄ and hydrolyze, the linker breaks to release the peptide in the form of an aldehyde.  The authors of this chapter wrote, "As described previously (54), owing to the presence of several amide functions, the equivalent amount of LiAlH₄ had to be increased with the length of the peptide."  I have not yet obtained the primary reference upon which this sentence is based, but I assume that it is a later paper.  One author of Chapter 6 is a coauthor of reference 54.  Reference 54 is "Synthesis of aldehydic peptides inhibiting renin" Fehrentz J-A, et al. Int. J. Peptide Protein Res. 26, 1985, 236-241.  PMID: 3902691  DOI: 10.1111/j.1399-3011.1985.tb03201.x
These authors used 8 equivalents of hydride (2 moles of LAH per mole of peptide) for a tripeptide, their compound 6. 

We have obtained good results with 1.5 equivalents Dibal-H for a BOC-protected amino acid-Weinreb amide; therefore, I was planning to start with this reducing agent.  What I am wondering is whether or not one could make an informed guess about the number of equivalents of Dibal-H that would be needed (offhand, I might guess 2.5 instead of 1.5).  The alternative is to make additions of this reagent empirically on the fly.

BTW, I recrystallized some dinitrophenyl hydrazine, which we use for a TLC dip.  It performs reasonably well for us, but DNP must be handled carefully. I have also heard of another reagent that detects aldehydes, but the name escapes me at the moment.