Determination whether the lack of functional AQP2 in kidneys of patients suffering from nephrogenic diabetes insipidus induces cyst formation

The Aquaporin-2 (AQP2) water channel is expressed in the renal collecting duct
and is essential in the vasopressin-regulated reabsorption of water from
pro-urine and concentration of urine. The essential role of AQP2 in this
process is illustrated by the fact that mutations in the AQP2 gene, leading to
non-functional water channels, is the cause of autosomal congenital nephrogenic
diabetes insipidus (cNDI)(1-5). In cNDI, the kidney is unable to concentrate
urine, despite increased vasopressin levels, resulting in voiding of large
volumes of low-osmotic urine (polyuria) and, as a consequence, polydipsia
(excessive thirst). In a recent publication in the renommated Journal of the
Americal Society of Nephrology (impact 9.0), evidence has been gathered leading
the authors to postulate that AQP2 is, besides preventing NDI, also essential
to prevent development of renal microcysts(6). The argumentation and evidence
the authors supply for this are that:
1) Inactivation of AQP2, AQP3 or AQP4 in transgenic knockout mice leads in all
cases to NDI, but that only mice with AQP2 mutations develop renal cysts.
2) Cells (LLCPK and MDCK), which are stably transfected with AQP2 show, during
cell proliferation, binding to integrin, a protein located in the basal matrix
and essential for cellular polarization. These authors suggest that this
binding is important for collecting duct cells to migrate in a particular
direction, which would be of relevance for tubule formation.

We have strong arguments that AQP2 is not involved in cell migration over the
basal matrix and that the lack of functional AQP2 does not lead to cyst
formation. Our arguments/indications are the following:
Ad 1) mice with congenital inactivating mutations in the AQP2 gene die within
1-2 days postnatally, because of severe dehydration, whereas mice with
congenital AQP1, AQP3 or AQP4 mutations survive well, but have NDI (7-11). The
lack of AQP2 thus gives a larger polyuria than the lack of AQP1, AQP3 or AQP4.
With humans, this difference in aquaresis is similar, but less severe: humans
with AQP2 mutations survive, but suffer from cNDI, whereas humans with AQP1 or
AQP3 mutations have no problem with their diuresis (12, 13).
Ad 2) When cells are polarized, they do not divide (proliferate) and
proliferating cells are thus not polarized. The cells used by Chen et al do
express AQP2 at all stages of cell proliferation, because they have been stably
transfected. From our analysis with mpkCCD mouse kidiney collecting duct cells,
which express endogenous AQP2 upon stimulation with vasopressin, it is clear
that AQP2 is only expressed in collecting duct cells when they are polarised,
but not when they are non-polarised, i.e. proliferative (see figure 1). Based
on our data, using a proper physiological cell model in contrast to Chen et al,
AQP2 is thus not available to bind integrin during proliferation and thus
cannot be involved in guided growth along the cellular matrix and tubule
formation.
Ad 3) In the >15 years that I have performed research on cNDI, I have never
encountered a clinician stating that he/she has observed cysts in the kidney of
a cNDI patient.
Our argumentation and data indicate that reduced polarization (and thus
increased proliferation) of collecting duct principal cells leads to reduced
expression of AQP2 and diuresis, but not that AQP2 itself is needed to prevent
cyst formation. The absence of microcysts in kidneys of cNDI patients with AQP2
gene mutations provides the answer.

Further fundament and clinical relevance:
Autosomal dominant polycystic kidney disease (ADPKD) is the most common genetic
kidney disease leading to chronic kidney failure and, eventually, lifelong
dialysis or, kidney transplantation. Similarly, nephronophthisis (NPHP) is the
most common cause of chronic kidney failure in children ADPKD is caused by
inactivating mutations in the polycystin-1(PC1) gene. Together with NPHP
proteins, they are involved in cilia formation. The first signs of development
of these diseases is increased diuresis, despite increased AVP levels. As ADPKD
and several NPHP forms mainly start to develop in the collecting duct, the
observed diuresis indicates to reduced AQP2 levels. Recently, it has been shown
that mutations in PC-1 lead to increased cell proliferation (14). Based on the
data from Chen above, the diminished AQP2 expression could contribute to the
cyst formation found in ADPKD (and NPHP). If our proposed analysis of the
kidney of NDI-AQP2 patients shows that the lack of AQP2 does NOT lead to cyst
formation, a contribution of AQP2 to cyst development in ADPKD and NPHP can be
excluded. Considering the severity of ADPKD and NPHP, this knowledge will not
only be of fundamental knowledge, but will also be of use in the development of
appropriate therapies for ADPKD and NPHP.

Why MRI instead of ultrasound?:
In sections of the kidneys of mice transgenic for AQP2 mutations, microcysts
were detected. Analysis of the kidneys of NDI patients need to be done
non-invasive. To be able to strongly establish whether kidneys of NDI patients
due to AQP2 mutations have microcysts, MRI is needed instead of ultrasound, as
this non-invasive technique has a resolution high enough to detect microcysts.

Selected population:
The only strong evidence whether AQP2 is involved in cyst formation is to
analyse whether humans lacking functional AQP2 have renal,cysts. Humans lacking
functional AQP2 are congenital NDI patients. Though this is a rare congenital
disease, the paediatrics dept of the Radboud Univ. Medical Center has been a
referral center for NDI patients for decades and has contacts with several such
patients.

Does the lack of the Aquaporin-2 water channel cause cyst formation in patients
suffering from nephrogenic diabetes insipidus?

This observational study will be carried out in the Radboud university medical
center. Since this study is only used to determine whether there is a presence
of microcysts and since autosomal congenital nephrogenic diabetes insipidus
with mutations in the AQP2 gene is quite unique, 2 to 5 patient will be asked
to participate. Eligible patients will be fully informed about the reasons
behind this observational study and so the need of a MRI. After having obtained
their written informed consent about participating in the study and the
availability of their kidney images for scientific publications, an MRI
appointment will be scheduled with at the radiology department.

At the radiology department, the MRI technicians will position the patient in
the MRI, supine with feet first and a body phase array coil to improve image
resolution. The MRI of the kidneys will be made without contrast agent. If
wished by the patient or relatives, the patient can practice the MRI with
pedagogical assistance upfront.
Since the kidneys have respiratory motion, the images will be made with fast
imaging techniques. If possible for the patient the scan will be performed
during one expiration because the kidney position is more constant in
expiration than in inspiration. If the sequence is too long to perform in one
breath-hold respiratory gating can be used. With this technique a navigator
pulse is used which monitors diaphragm motion to trigger when images can be
made. A phased array body coil is used to improve the signal-to-noise ratio
since high contrast is important in visualizing microcysts, a high
signal-to-noise ratio is essential. The patients arms should be raised above
the head to prevent aliasing in coronal imaging. Images will be made with a 1,5
or 3 Tesla system.
The following sequences will be made:
- Coronal T2-weighted sequence, serving as a localizer, but also supplying
valuable T2-weighted information. The limitation of this sequence is a
relatively low signal-to noise ratio.
- Axial T2-weighted turbo spin echo sequence with fat suppression to provide
more detailed T2-weighted information. The T2-weighted sequence is especially
helpful in characterizing cysts.
- Axial T1-weighted gradient echo sequence preferably as a dual-echo sequence.
Many solid renal lesions, like cysts, are hypointense compared to the renal
parenchyma on T1-weighted images, but lesions with high protein content may
show hyperintense signal.

The MRI will take about 30 minutes after it is made, the radiologist will
interpret the obtained images and make a report. This report will be send to
the researchers involved and general practitioner of the patient as well.

Adverse events that can be expected are primarily related to the process
acquiring the MRI.
These events are for instance; claustrophobia, dizziness, and a local very
minimal rise in temperature of the upper surface of the body (skin).

This study is fundamental in nature and will not have clear benefits for the
participant.
When the MRI is made, more could be known about the condition of the patient*s
kidneys. There is as well a small change of unexpected findings.
A MRI is a standard imaging technique which has a relatively low discomfort for
the patient.
During the MRI patients can get scared because of the loud noise of the machine
or the narrow space in which they are lying this could give problems when a MRI
needs to be made in the future.