The so-called Markina Gora skeleton from Kostenki in Russia (near the Don River) was tested for mtDNA and determined to belong to haplogroup U2. This haplogroup is found at very low frequencies in southern Russia nowadays (actually its subclade U2e), but is otherwise primarily found in and around India, Pakistan, Tajikistan and Kyrgyzstan, and to a lower extent Iran and Xinjiang.
If people carrying hg U2 lived in Russia 30,000 years ago, it is doubtful that they already lived all over South Asia. The possibility of a South Asian hunter-gatherer travelling on foot all the way to European Russia is far-fetched. U2 is more probably a Paleolithic European subclade, like U4 and U5. Its widespread presence in the Indian subcontinent today can be explained in the same way that Y-haplogroup R1a is likewise widespread there : the Indo-European migrations.
I had a doubt, however, when I saw the facial reconstruction of the skeleton, which doesn't appear particularly European. But traits do change over such a long period of time, especially since modern European are ethnically mixed compared to Paleolithic people.
The little of U2 left in Europe is found almost exactly in the presumed Indo-European homeland, i.e. between the Dniester and Ural Rivers, in the Pontic-Caspian steppe north of the Black Sea, Caucasus and Caspian Sea. If anything U2 fits better the IE homeland than R1a. U2 is not present in any representative level in Kazakhstan (and R1a is also weak there compared to surrounding countries).
What bothers me with this theory is that the 4000 years since the IE migrations is a very short time to develop the numerous subclades of U2 existing today in South Asia. The U2e found in Russia today is itself a branch of U2 shared by U2c and U2d (all three have the 152 mutation as you can is on the mtDNA tree). We could imagine that the main subclades of U2 developed in the Pontic-Caspian steppes before the IE migrations, and that all of the U2 women except a few U2e ones left for South Asia en masse, without leaving any trace behind them. That would be an exceptional event in history, since there always seem to be a trail left behind when it comes to DNA. Why would other mt-haplogroups like H6, U4 or W be present both in Europe and South Asia, but U2 move completely and thoroughly to South Asia apart from a tiny fraction of U2e ? This is very perplexing.
What Y-haplogroup could the Markina Gora skeleton be ?
Naturally, it would be wrong to assume that this 30,000 year-old hunter-gatherer from Kostenki had an R1a Y-chromosome - simply because R1a did not exist yet (R* had barely appeared, but probably in Central Asia). Unfortunately his Y-DNA wasn't tested, but it's hard to conceive that he would have been anything else but IJ or I*, or an extinct haplogroup not listed in the Y-tree.
R1a people would have come a few thousands years later, maybe as R1* and replaced most of the paternal lineages in the Eurasian steppe.
I am not entirely ruling out that R1 or R1a could already have existed in the Eurasian steppe 30,000 years ago, but that would be strongly at odd with the calculations made by specialists at ISOGG and elsewhere regarding the age of R1a.
For once the BBC has a decent article on the "discovery" of U2 (they even mention the haplogroup, which is new for them), with a good summary for beginners of the technicalities of ancient DNA testing.