The infamous laryngeal theory states that Sanskrit a originates from PIE e/o, but this is a faulty statement, showing how little the theory's supporters know about so called laryngeals or semetic languages. The a coloring laryngeal, as it is called, is said to have colored PIE e to a in Sanskrit, but from what we know the opposite would've occurred. And what better way to prove that then to go to Arabic, a language infested with laryngeal consonants. In Arabic, a is advanced to æ in the environment of most consonants:
- (/m/, /b/ and /f/),
- plain (non-emphatic) with the exception of /r/ (/θ/, /ð/, /n/, /t/, /d/, /s/, /z/, /l/, /ʃ/ and /d͡ʒ~ɡ~ʒ/)
- (/ħ/ and /ʕ/)
- (/h/ and /ʔ/)
- /j/, /k/ and /w/;
- Across and West Asia, the allophones [æ] and [ɑ] may be realized differently, either as [ ~ ~ ], or both as [ ~ ]
- In northwestern Africa, the open front vowel /æ/ is raised to ɛ or e.
But lets not stop at Arabic. It is well known that Proto Semitic ā became ō in Ancient Hebrew, for reasons not fully understood. Changes like this also occurred in the Great Vowel Shift, where long a become e. The opposite shift however, was not observed.
In Romani words derived from Sanskrit, the change was common. See this link for more information.
I conclude from these facts that the Sanskrit vowel a is the original, which shifted to æ, then to e in languages like Greek and Latin. In some places it shifted to o, this shift being less common than the former one. This reconstruction is typologically stable, given that a plain unrounded open vowel is the most common sound in the world. Such dubious claims of PIE having had only /e/ are to be regarded warily. Abkhaz has only two distinct vowels (a/ə) , not e, a fact that has so far been ignored.
Though some will look the Hittite for objections, the fact remains that Hittite ḫ and ḫḫ almost never occur initially, something that should be expected of any laryngeal, and are normally followed by an a. These sounds would have been e and o coloring laryngeals if pharyngeal, as these consonants tend to draw following a towards æ, and if labial, towards o.
Here is an enlightening article that deals with the subject. http://new-indology.blogspot.com/2013/07/indo-european-linguistics-indo-iranian.html