Hello, Salmonella typhimurium doesn't seldom cause enterocolitis in mice. Through its invasion proteins, S. typhimurium targets the gut associated lymphoid tissue to produce typhoid-like symptoms (which happens to make them good models for that kind of disease). The theory behind the pre-treatment with streptomycin is to deplete the naturally occurring microbiota that protects the host and hinders the ability of Salmonella  to colonize. The pre-treatment with sodium bicarbonate is to neutralize the stomach's pH so a higher bacterial load can make its way to the intestine. If you want to study salmonellosis in the gut, you will need to treat with streptomycin.  Attached to this post is a paper from a leading group on gastrointestinal pathogenesis in Canada, their methods have been validated multiple times and it works really well. Hope this helped :) 

http://iai.asm.org/content/82/9/3753.long

Thank you Guillaume. I want to use S. typhimurium infection model to study dendritic cells in gut, not salmonellosis. Then, it doesn't matter that I pre-treat mice with sodium bicarbonate, right? The bacteria strain I use is ATCC 14028, I'm afraid it isn't resistant to streptomycin.

Oh sorry about that I misunderstood :). If you want to study dendritic cell in the context of intestinal inflammation caused by Salmonella, you'll need to pre-treat the mice with streptomycin. You use streptomycin to clear the microbiota before infecting with Salmonella (around 24 hours before). Since you stop the streptomycin before adding the bacteria, the antibiotic should not be present in the intestine anymore so the resistance is not required for the experiment ! I hope this answers your question :)

Also, the pre-treatment with sodium carbonate should not be required if you're using mice that are susceptible to the bacteria. For example, C57Bl6 are highly susceptible to Salmonella infections because they are Nramp1 -/-, you might need to pre-treat mice that are more resistant to the bacteria to make sure that an infectious dose makes its way to the intestine !

Thank you very much, Guillaume. Actually, I 've already crossed a kind of transgenic mice whose background is B57BL/6, so it's hard for me to use another kind of mice...

Maybe I can reduce infection dose ! :)

You can make 14028s strain resistant to streptomycin by using P22 phage to transduce the aadA gene from strain SL1344. In this way you would have a StrR strain without the drawbacks of a ribosomal mutation.