Lets not forget assimilatory nitrate reduction. If carbon is available and ammonium is not the microbial population will use nitrate. Also if we include plants in the soil ecosystem, assimilatory reduction is the dominant fate of nitrate.

Nitrate can be biologically reduced by two well-known pathways-

1. Denitrification (reduction of nitrate to nitrogen via nitrite)

2. DNRA (dissimilatory nitrate reduction to ammonia)

The former is more common and may occur even in the presence of some oxygen. The later occurs under strict anoxic conditions. 

If DNRA is active in the soil, then it might result in the accumulation of ammonium, which can be used for biological metabolic pathways for organic matter synthesis. 

The DNRA pathway is not common in soil. If anoxic conditions do occur, this will drive denitrification process which means in soil, denitrification process is likely to dominate DNRA process.

In some cases, with high presence of NO2- (nitrite), NH4 could be converted into N2, a process referred to as anammox. This process also requires anoxic conditions.

I am interested to know whether there are documented evidence of NO3 being directly absorbed and synthesised (intra-cell) by soil bacteria to form protein, similar to heterotrophs directly absorbing NH4 and synthesise into protein.

Lets not forget assimilatory nitrate reduction. If carbon is available and ammonium is not the microbial population will use nitrate. Also if we include plants in the soil ecosystem, assimilatory reduction is the dominant fate of nitrate.

Soil is not a homogenous medium. The spatial variability of soil will invariably mean that nitrate will be immobilized in the presence of ammonium. There may be organisms that are unable to utilize ammonium... but I am unaware of them and it seems unlikely that they would be very common.

This is not a direct answer, but many fungi will grow happily on lab media with nitrate as the sole source of N (e.g., Czapek-Dox agar). The N must be converted to microbial protein. However, growth conditions are aerobic and the medium also contains readily available C so I'm not sure how readily the culture information translates to poorly aerated soils. Very few bacteria grow on C-D agar.

All answers have attempted to assist and thanks. However, to move forward, I suggest we focus on the assimilatory nitrate reduction aided by soil bacteria.

While N15 is an excellent tool, it is difficult to infer from many such studies whether there was direct NO3 assimilation by soil bacteria. If labelled NO3 is used, specific attention is required to ensure it has not transformed into NH4 (or at least identify this) before being assimilated by bacteria. What makes NO3 'immobilsation' studies difficult and complex is that NO3 is more dynamic than NH4. Unlike NH4, nitrate is susceptible to denitrification and dissimilatory reduction (both of which requires dissolved-C or reliable electron donor) and leaching.

Ironically, bacterial or fungal immobilisation does require C which is also required for denitrification and DNRA.

Once NO3 is absorbed by bacteria, it has to be reduced to be able to be assimilated, which is considered as an energy 'taxing' process. Therefore, if both NH4 and NO3 are present it could be expected bacteria will prefer NH4 over NO3. Judging by Karolina's answer, NO3 was assimilated in the presence of NH4. The question arises whether, on that occasion NO3 went through a DNRA process before being assimilated by bacteria or directly absorbed and assimilated?

If bacteria can take up NO3 directly, the question is (1) can all heterotrophs assimilate NO3 in the absence of NH4? and (2) in the presence of NH4 which process drives bacterial uptake of NO3?

There is a quite old now but relevant literature about this question, both in soils and with isolated bacteria. As stated before in this discussion, microorganisms are able to use both NH4 and NO3 sources, but the use of NO3 sources costs more, due to the step of reduction of nitrate to ammonium before ammonium assimilation. It has been shown that when both sources are available the microorganisms are using preferentially NH4 sources (lower cost), but this is quite theoretical, because in the real conditions of soils the availability of inorganic N is heterogeneous, and in a given volume of soil, there are sites where the two forms are not available, so both processes may occur. You could have a look to a paper from our group (Recous et al. 1990 published in Soil Biology and Biochemistry) with results on this topic but also a range of references from the 50 - 80' in microbiology and soil science on this question. 

Thanks for the comments Sylvie. The timing of your comments is perfect since I have just spent my weekend time sifting through a large number of papers on this topic. From an extensive literature review, I draw the following conclusions:

(1) Best way to tackle the question on direct NO3 assimilation is to deal with bacterial metabolisms. The key is that cytoplasmic assimilatory nitrate reductase (Nas) is responsible for the direct assimilation of NO3. This enzyme is induced by NO3 and repressed by NH4 but not affected by O2. As pointed out by Sylvie, direct assimilatory process is very energy taxing which includes the use of ATP to transport NO3 into the cell with the subsequent assimilation demanding 8 electrons to convert to NH4.

(2) Phyllogenetically, the above enzyme (Nas) is widespread in bacteria and archaeal heterotrophs. Based on this evidence alone it could be argued the potential for direct assimilation of NO3 is high. However, before we postulate or conclude on this basis we need to consider more common metabolic processes as well.

(3) A commonly found and widely documented metabolic process denitrification is a respiratory process induced by nitrate (used as an effective electron acceptor with electron donors being NADH, ferredoxin and flavodoxin) to created ATP (Nar system) is carried out under anaerobic conditions hence repressed by O2 but insensitive to NH4. I am convinced of this process and hence not debating this process.

(4) Often called fermentative or DNRA process, the periplasmic dissimilatory NO3 is not repressed by O2 or NH4. This metabolic process is simply a redox balancing exercise. Compared to the numerous studies on N-immobilisation, DNRA studies have received relatively low attention from soil scientists. However, this process is being increasingly researched and documented lately but as a stand-alone study.

(5) From my extensive review, I do not have any current doubt about the ability for soil bacteria to directly assimilate NO3. However, judging by other dominant and low energy demanding metabolic processes (e.g. DNRA) in soil and in the absence of DNRA consideration in many 15NO3 bacterial assimilation or immobilisation studies I am yet to be convinced that direct assimilation does occur in the presence of NH4. I do agree with Karolina and David that NH4 is less mobile than NO3 and hence in soil microsites the only bacterial N source may well be NO3. However, if such a situation does exist the question is whether DNRA process takes effect before direct assimilation of NO3.

(6) Many research papers on NO3 immobilisation, including 15N works have failed to account for the DNRA process. It may be a difficult process to isolate in a dynamic soil bacterial/fungal metabolic environment. Judging by the high bacterial preference to NH4 there may be no opportunity for NH4 to accumulate by a DNRA process, particularly in an NH4 limiting microsite environment. Therefore monitoring NH4 chemically with time may not yield much.

(7) Despite the above difficulty I recommend DNRA be one of the key components of every 15NO3 direct assimilation soil studies.

I agree Karolina and we could add temperature and moisture content as other key factors controlling the general bacterial assimilation processes. This also raises another interesting point of mineralisation-immobilisation turnover of N. While the net immobilisation-mineralisation process is driven by CNP ratios, both immobilisation and mineralisation processes do occur simultaneously. While immobilisation is mainly a biosynthetic process (anabolic), mineralisation is either a respiratory or a fermentative (both catabolic) process. While ammonium can be used for biosynthetic process  or as an electron donor in the nitrification process, it is unable to take up the electron acceptor role like NO3 in a respiratory process. Owing to its second best O2 status in soils, nitrate's predominant role in bacterial respiratory processes in soil potentially renders it a low probability direct bacterial protein base.