dear Julie, 

My opinion is that stress in its proper sense involves the activation of the HPA axis. 

Stress is defined as a state of real or perceived threat to homeostasis, which maintenance in the presence of stressors requires activation of  responses involving the endocrine, nervous, and immune systems. 

Dialogues Clin Neurosci. 2006 Dec; 8(4): 383–395.The role of the hypothalamic-pituitary-adrenal axis in neuroendocrine responses to stress

Sean M. Smith, Wylie W. Vale,

I agree with Beatrice and Renzo, stress in itself is defined as anything that activates the HPA axis, and I'm not sure that an organism would even be viable without an HPA axis (obviously not all organisms have an HPA axis, but they do have analogous homeostatic systems) unless it was a parasite.

The HPA axis regulates all kinds of homeostatic processes like digestion, immune responses, and energy metabolism. Even a small dysregulation in its function can have devastating effects--disorders like PTSD and depression involve exaggerated or diminished negative feedback of the HPA axis. But to not have a functioning HPA axis at all would end all functioning.

But for the sake of discussion, in an experimental model without an HPA axis, there would still be the other systems involved in responses to stress--the sympathetic nervous system, brainstem responses, frontal cortex and limbic system (assuming we're talking about higher-order organisms) responses can all be directly "activated" by external stress.

Somewhat related, here is a fascinating story from a person who had part of her pituitary removed and really could not function without hormone replacement http://www.nytimes.com/2013/11/24/opinion/sunday/bringing-back-my-real-self-with-hormones.html?_r=0

Hope that helps!

As I understand, in a Neuroscientific view, stress is a multifactorial phenomenon, involving a disbalance in the HP, as well as in the vagal, cardiovascular, immunological, behavioral, emotional and psychological fields, just to mention the most proeminent.

Dear Julie, there are several facets to the binomium stress - HPA axis. We have to consider the stressful stimulus, which may be modulated by a number of factors that may result in attenuation of the HPA axis activation and glucocorticoids secretion. Attached, is a review paper that illustrates this modulation and proposes a new conceptualization of stress. I believe it is worth reading. Best of luck with your work

For sure, glucocorticoids (cortisol and/or corticosterone; depending upon the species) are the end products of the HPA axis and named "stress hormones". And the HPA axis plays a central (key!) role in stressful situations. And the neuroendocrine axis (HPA axis) is sometimes even labelled as THE "stress response". 

BUT that's not 100% correct: As indicated above, stress is a multifactoral phenomenon and there are more stress responses for an individual available (also behaviour and the immune system can act directly). But, most important, there's also the autonomic nervous system (sympatho-adrenomedullar axis) which can be activated in response to stressor. This system is the basis of Cannon's proposed "fight and flight" response which result from an activation of the adrenal medulla and the realease of catecholamines (adrenaline or epinephrine, as the main stress hormone of this axis is named in Europe or the US, respectively). This axis is harder to monitor (mainly done via the biological effects of catecholamines (measured as heart rate variability), quicker turned on and of shorter duration.  

I fully agree with Renzo and the others. I'd like to add that the autonomic nervous system is also an important part of the equation, and stress was first studied in relation to catecholamine concentration (e.g., Cannon) before cortisol was considered.

There is not a 1:1 relationship between stress and cortisol. The best demonstration of that is in the responses of phobic patients to their feared stimulus  in exposure therapy. HPA axis activation is not seen although they are markedly aroused and anxious during the procedure. Run a search for papers authored by George C. Curtis in the 1970s and 1980s.

If you use  a model that has no HPA axis, I think it might be a bit problematic to present it as a stressed model later. The HPA axis activation is , by definition, needed for naming a manipulation "stress(or)". 

Dear Juli,

I am pleased that our colleagues follow (as well as me) the original idea of Hans Selye according to which signs of suprarenal gland activation (cortisol production) always         take part in stress reaction development. I believe it is like that because of necessity of sufficiently high glycaemia during the whole duration of the stress which cannot be predicted.  Considering participation of only sympathetic nerve system in stress reaction, limiting condition for stress survival would be actual individual store of glycogen. Gluconeogenesis must be started sufficiently early. As always it is about the bioenergetics… regards mummy

Excellent question. I believe I understand the question you pose here, I shall attempt to rephrase it here. So please tell me if I am correct:

Is there a way to produce a stress-like effect in experimental animals with an absent or dysfunctional HPA axis?

I do not know the answer to this, but I do appreciate the depth of your question. It is a valid and educated question, [especially] when considering that biological organisms seem to have evolved multiple paths to attain overlapping outcomes. Consider overlaps between:

1. Immunoglobulin and complement factors in opsinization

2. Intrinsic and extrinsic clotting pathways

3. Vitamin D and parathyroid hormones in calcium homeostasis

4. e.t.c.

So this is definitely an excellent question. You must be good at pattern recognition. 

Hi Bianchi. I am assuming by "these processes" you are referring to those mentioned in my response. If not, then ignore the discussion below.

If we approach the stated processes as interconnected, and also agree that we have come to such conclusion from experimental verification; how then can we conclude that such connections do not exist for the physiology of stress, if no experimental results have been produced in support or refutation of the stated argument. In other words, the question posed by Bienertova-Vasku is very educated and can only be answered by an experiment involving observations and measurements of the right set of parameters. The question then will be to choose the right set of parameters that seem to be consistent in all forms of stress, and can therefore be applied as (at least) a minimum definition of stress.